commit 93e014f831c0e662769eae693119811a14843491
Author: aboelhamd <aboelhamd.abotreka@gmail.com>
Date:   Tue Jul 23 12:54:57 2019 +0200

    Finished training all datasets

diff --git a/all-datasets-models.ipynb b/all-datasets-models.ipynb
new file mode 100644
index 0000000..48d893e
--- /dev/null
+++ b/all-datasets-models.ipynb
@@ -0,0 +1,15585 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": 1,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "%matplotlib inline"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "\n",
+    "# Classifier comparison\n",
+    "\n",
+    "\n",
+    "A comparison of a several classifiers in scikit-learn on synthetic datasets.\n",
+    "The point of this example is to illustrate the nature of decision boundaries\n",
+    "of different classifiers.\n",
+    "This should be taken with a grain of salt, as the intuition conveyed by\n",
+    "these examples does not necessarily carry over to real datasets.\n",
+    "\n",
+    "Particularly in high-dimensional spaces, data can more easily be separated\n",
+    "linearly and the simplicity of classifiers such as naive Bayes and linear SVMs\n",
+    "might lead to better generalization than is achieved by other classifiers.\n",
+    "\n",
+    "The plots show training points in solid colors and testing points\n",
+    "semi-transparent. The lower right shows the classification accuracy on the test\n",
+    "set.\n",
+    "\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 2,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import os\n",
+    "\n",
+    "path='sklearn-nobad'\n",
+    "files = {}\n",
+    "# r=root, d=directories, f=files\n",
+    "for r, d, f in os.walk(path):\n",
+    "    for file in f:\n",
+    "        files[file]=os.path.join(r, file)\n",
+    "\n",
+    "os.mkdir('models')"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 4,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "file name : 40+68_63+.csv\n",
+      "Rules(classes) number : 2\n",
+      "Words(features) number : 2\n",
+      "Records number : 276"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.5</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>ser</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.5</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>ser</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule  weight  word1 word2\n",
+       "0     0     0.5  haber   ser\n",
+       "1     1     0.5  haber   ser"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.45654482962107334\n",
+      "model : LinearSVM , score = 0.4637975614023888\n",
+      "model : RBFSVM , score = 0.31829418475530613\n",
+      "model : DecisionTree , score = 0.3836062964213818\n",
+      "model : RandomForest , score = 0.38350158147992314\n",
+      "model : AdaBoost , score = 0.3617831020952113\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 27+21_106_29+10_30_106_29+10_110_29_106_29+11_30_106_29+11_110_29_106_29+33_1_30_106_29+33_1_110_29_106_29+.csv\n",
+      "Rules(classes) number : 8\n",
+      "Words(features) number : 6\n",
+      "Records number : 120"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "      <th>word5</th>\n",
+       "      <th>word6</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.137765</td>\n",
+       "      <td>uno</td>\n",
+       "      <td>aportación</td>\n",
+       "      <td>muy</td>\n",
+       "      <td>claro</td>\n",
+       "      <td>y</td>\n",
+       "      <td>positivo</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.128679</td>\n",
+       "      <td>uno</td>\n",
+       "      <td>aportación</td>\n",
+       "      <td>muy</td>\n",
+       "      <td>claro</td>\n",
+       "      <td>y</td>\n",
+       "      <td>positivo</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.119732</td>\n",
+       "      <td>uno</td>\n",
+       "      <td>aportación</td>\n",
+       "      <td>muy</td>\n",
+       "      <td>claro</td>\n",
+       "      <td>y</td>\n",
+       "      <td>positivo</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>0.119732</td>\n",
+       "      <td>uno</td>\n",
+       "      <td>aportación</td>\n",
+       "      <td>muy</td>\n",
+       "      <td>claro</td>\n",
+       "      <td>y</td>\n",
+       "      <td>positivo</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>4</td>\n",
+       "      <td>0.127315</td>\n",
+       "      <td>uno</td>\n",
+       "      <td>aportación</td>\n",
+       "      <td>muy</td>\n",
+       "      <td>claro</td>\n",
+       "      <td>y</td>\n",
+       "      <td>positivo</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>5</th>\n",
+       "      <td>5</td>\n",
+       "      <td>0.127315</td>\n",
+       "      <td>uno</td>\n",
+       "      <td>aportación</td>\n",
+       "      <td>muy</td>\n",
+       "      <td>claro</td>\n",
+       "      <td>y</td>\n",
+       "      <td>positivo</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>6</th>\n",
+       "      <td>6</td>\n",
+       "      <td>0.119732</td>\n",
+       "      <td>uno</td>\n",
+       "      <td>aportación</td>\n",
+       "      <td>muy</td>\n",
+       "      <td>claro</td>\n",
+       "      <td>y</td>\n",
+       "      <td>positivo</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7</th>\n",
+       "      <td>7</td>\n",
+       "      <td>0.119732</td>\n",
+       "      <td>uno</td>\n",
+       "      <td>aportación</td>\n",
+       "      <td>muy</td>\n",
+       "      <td>claro</td>\n",
+       "      <td>y</td>\n",
+       "      <td>positivo</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight word1       word2 word3  word4 word5     word6\n",
+       "0     0  0.137765   uno  aportación   muy  claro     y  positivo\n",
+       "1     1  0.128679   uno  aportación   muy  claro     y  positivo\n",
+       "2     2  0.119732   uno  aportación   muy  claro     y  positivo\n",
+       "3     3  0.119732   uno  aportación   muy  claro     y  positivo\n",
+       "4     4  0.127315   uno  aportación   muy  claro     y  positivo\n",
+       "5     5  0.127315   uno  aportación   muy  claro     y  positivo\n",
+       "6     6  0.119732   uno  aportación   muy  claro     y  positivo\n",
+       "7     7  0.119732   uno  aportación   muy  claro     y  positivo"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.06703474247602752\n",
+      "model : LinearSVM , score = 0.10093532974509636\n",
+      "model : RBFSVM , score = 0.11605573183596676\n",
+      "model : DecisionTree , score = 0.0\n",
+      "model : RandomForest , score = 0.0\n",
+      "model : AdaBoost , score = 0.033775670441412345\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 13+37_12+37_113_1+.csv\n",
+      "Rules(classes) number : 3\n",
+      "Words(features) number : 3\n",
+      "Records number : 117"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.352679</td>\n",
+       "      <td>hacer</td>\n",
+       "      <td>50</td>\n",
+       "      <td>año</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.323660</td>\n",
+       "      <td>hacer</td>\n",
+       "      <td>50</td>\n",
+       "      <td>año</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.323660</td>\n",
+       "      <td>hacer</td>\n",
+       "      <td>50</td>\n",
+       "      <td>año</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight  word1 word2 word3\n",
+       "0     0  0.352679  hacer    50   año\n",
+       "1     1  0.323660  hacer    50   año\n",
+       "2     2  0.323660  hacer    50   año"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.2942577385293566\n",
+      "model : LinearSVM , score = 0.3158541827161679\n",
+      "model : RBFSVM , score = 0.3532454955917427\n",
+      "model : DecisionTree , score = 0.17141174651482455\n",
+      "model : RandomForest , score = 0.1712896936630942\n",
+      "model : AdaBoost , score = 0.1694033114534203\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 76_64+76_91+76_108_68+75_67+75_63_64+75_63_91+75_63_108_68+100_40_64+100_40_91+100_40_108_68+100_37_67+100_37_63_64+100_37_63_91+100_37_63_108_68+.csv\n",
+      "Rules(classes) number : 14\n",
+      "Words(features) number : 5\n",
+      "Records number : 56"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "      <th>word5</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.073716</td>\n",
+       "      <td>no</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>hablar</td>\n",
+       "      <td>de</td>\n",
+       "      <td>emplear</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.071830</td>\n",
+       "      <td>no</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>hablar</td>\n",
+       "      <td>de</td>\n",
+       "      <td>emplear</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.070473</td>\n",
+       "      <td>no</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>hablar</td>\n",
+       "      <td>de</td>\n",
+       "      <td>emplear</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>0.070710</td>\n",
+       "      <td>no</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>hablar</td>\n",
+       "      <td>de</td>\n",
+       "      <td>emplear</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>4</td>\n",
+       "      <td>0.070710</td>\n",
+       "      <td>no</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>hablar</td>\n",
+       "      <td>de</td>\n",
+       "      <td>emplear</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>5</th>\n",
+       "      <td>5</td>\n",
+       "      <td>0.069735</td>\n",
+       "      <td>no</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>hablar</td>\n",
+       "      <td>de</td>\n",
+       "      <td>emplear</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>6</th>\n",
+       "      <td>6</td>\n",
+       "      <td>0.070710</td>\n",
+       "      <td>no</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>hablar</td>\n",
+       "      <td>de</td>\n",
+       "      <td>emplear</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7</th>\n",
+       "      <td>7</td>\n",
+       "      <td>0.072360</td>\n",
+       "      <td>no</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>hablar</td>\n",
+       "      <td>de</td>\n",
+       "      <td>emplear</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>8</th>\n",
+       "      <td>8</td>\n",
+       "      <td>0.071338</td>\n",
+       "      <td>no</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>hablar</td>\n",
+       "      <td>de</td>\n",
+       "      <td>emplear</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>9</th>\n",
+       "      <td>9</td>\n",
+       "      <td>0.070000</td>\n",
+       "      <td>no</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>hablar</td>\n",
+       "      <td>de</td>\n",
+       "      <td>emplear</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>10</th>\n",
+       "      <td>10</td>\n",
+       "      <td>0.072360</td>\n",
+       "      <td>no</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>hablar</td>\n",
+       "      <td>de</td>\n",
+       "      <td>emplear</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>11</th>\n",
+       "      <td>11</td>\n",
+       "      <td>0.072360</td>\n",
+       "      <td>no</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>hablar</td>\n",
+       "      <td>de</td>\n",
+       "      <td>emplear</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>12</th>\n",
+       "      <td>12</td>\n",
+       "      <td>0.071338</td>\n",
+       "      <td>no</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>hablar</td>\n",
+       "      <td>de</td>\n",
+       "      <td>emplear</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>13</th>\n",
+       "      <td>13</td>\n",
+       "      <td>0.072360</td>\n",
+       "      <td>no</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>hablar</td>\n",
+       "      <td>de</td>\n",
+       "      <td>emplear</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "    rule    weight word1  word2   word3 word4    word5\n",
+       "0      0  0.073716    no  haber  hablar    de  emplear\n",
+       "1      1  0.071830    no  haber  hablar    de  emplear\n",
+       "2      2  0.070473    no  haber  hablar    de  emplear\n",
+       "3      3  0.070710    no  haber  hablar    de  emplear\n",
+       "4      4  0.070710    no  haber  hablar    de  emplear\n",
+       "5      5  0.069735    no  haber  hablar    de  emplear\n",
+       "6      6  0.070710    no  haber  hablar    de  emplear\n",
+       "7      7  0.072360    no  haber  hablar    de  emplear\n",
+       "8      8  0.071338    no  haber  hablar    de  emplear\n",
+       "9      9  0.070000    no  haber  hablar    de  emplear\n",
+       "10    10  0.072360    no  haber  hablar    de  emplear\n",
+       "11    11  0.072360    no  haber  hablar    de  emplear\n",
+       "12    12  0.071338    no  haber  hablar    de  emplear\n",
+       "13    13  0.072360    no  haber  hablar    de  emplear"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.0\n",
+      "model : LinearSVM , score = 0.07381804243862623\n",
+      "model : RBFSVM , score = 0.07381804243862623\n",
+      "model : DecisionTree , score = 0.0\n",
+      "model : RandomForest , score = 0.0\n",
+      "model : AdaBoost , score = 0.0\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 67+63_64+63_91+63_108_68+.csv\n",
+      "Rules(classes) number : 4\n",
+      "Words(features) number : 3\n",
+      "Records number : 116"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.254216</td>\n",
+       "      <td>encargar</td>\n",
+       "      <td>de</td>\n",
+       "      <td>elaborar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.254216</td>\n",
+       "      <td>encargar</td>\n",
+       "      <td>de</td>\n",
+       "      <td>elaborar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.247765</td>\n",
+       "      <td>encargar</td>\n",
+       "      <td>de</td>\n",
+       "      <td>elaborar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>0.243803</td>\n",
+       "      <td>encargar</td>\n",
+       "      <td>de</td>\n",
+       "      <td>elaborar</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight     word1 word2     word3\n",
+       "0     0  0.254216  encargar    de  elaborar\n",
+       "1     1  0.254216  encargar    de  elaborar\n",
+       "2     2  0.247765  encargar    de  elaborar\n",
+       "3     3  0.243803  encargar    de  elaborar"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.13992349932223666\n",
+      "model : LinearSVM , score = 0.19246829395922085\n",
+      "model : RBFSVM , score = 0.237588422251646\n",
+      "model : DecisionTree , score = 0.07014294386464993\n",
+      "model : RandomForest , score = 0.051613671230843566\n",
+      "model : AdaBoost , score = 0.12037676753559738\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 65+64_63+108_40+108_68_63+.csv\n",
+      "Rules(classes) number : 4\n",
+      "Words(features) number : 3\n",
+      "Records number : 252"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.25472</td>\n",
+       "      <td>por</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>adoptar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.25472</td>\n",
+       "      <td>por</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>adoptar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.24528</td>\n",
+       "      <td>por</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>adoptar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>0.24528</td>\n",
+       "      <td>por</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>adoptar</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule   weight word1  word2    word3\n",
+       "0     0  0.25472   por  haber  adoptar\n",
+       "1     1  0.25472   por  haber  adoptar\n",
+       "2     2  0.24528   por  haber  adoptar\n",
+       "3     3  0.24528   por  haber  adoptar"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.1854522284236849\n",
+      "model : LinearSVM , score = 0.20280748827570969\n",
+      "model : RBFSVM , score = 0.2596570585636813\n",
+      "model : DecisionTree , score = 0.12557739997430464\n",
+      "model : RandomForest , score = 0.09485241605291463\n",
+      "model : AdaBoost , score = 0.1167580398440067\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 7+8+9+10_108_1+11_108_1+33_1_108_1+.csv\n",
+      "Rules(classes) number : 6\n",
+      "Words(features) number : 4\n",
+      "Records number : 48480"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.150565</td>\n",
+       "      <td>el</td>\n",
+       "      <td>periodo</td>\n",
+       "      <td>de</td>\n",
+       "      <td>sesión</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.201926</td>\n",
+       "      <td>el</td>\n",
+       "      <td>periodo</td>\n",
+       "      <td>de</td>\n",
+       "      <td>sesión</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.206313</td>\n",
+       "      <td>el</td>\n",
+       "      <td>periodo</td>\n",
+       "      <td>de</td>\n",
+       "      <td>sesión</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>0.150565</td>\n",
+       "      <td>el</td>\n",
+       "      <td>periodo</td>\n",
+       "      <td>de</td>\n",
+       "      <td>sesión</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>4</td>\n",
+       "      <td>0.140065</td>\n",
+       "      <td>el</td>\n",
+       "      <td>periodo</td>\n",
+       "      <td>de</td>\n",
+       "      <td>sesión</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>5</th>\n",
+       "      <td>5</td>\n",
+       "      <td>0.150565</td>\n",
+       "      <td>el</td>\n",
+       "      <td>periodo</td>\n",
+       "      <td>de</td>\n",
+       "      <td>sesión</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight word1    word2 word3   word4\n",
+       "0     0  0.150565    el  periodo    de  sesión\n",
+       "1     1  0.201926    el  periodo    de  sesión\n",
+       "2     2  0.206313    el  periodo    de  sesión\n",
+       "3     3  0.150565    el  periodo    de  sesión\n",
+       "4     4  0.140065    el  periodo    de  sesión\n",
+       "5     5  0.150565    el  periodo    de  sesión"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.16419735349154174\n",
+      "model : LinearSVM , score = 0.1661864935934602\n",
+      "model : RBFSVM , score = 0.1609621342969151\n",
+      "model : DecisionTree , score = 0.15462650678497822\n",
+      "model : RandomForest , score = 0.15304989039368436\n",
+      "model : AdaBoost , score = 0.14956443099905853\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 25+26_1+16_106_29_1+10_29_106_29_1+11_29_106_29_1+33_23+33_24_1+33_14_106_29_1+33_1_29_106_29_1+.csv\n",
+      "Rules(classes) number : 9\n",
+      "Words(features) number : 6\n",
+      "Records number : 72"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "      <th>word5</th>\n",
+       "      <th>word6</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.118151</td>\n",
+       "      <td>el</td>\n",
+       "      <td>comisión</td>\n",
+       "      <td>europeo</td>\n",
+       "      <td>y</td>\n",
+       "      <td>vario</td>\n",
+       "      <td>estado</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.105356</td>\n",
+       "      <td>el</td>\n",
+       "      <td>comisión</td>\n",
+       "      <td>europeo</td>\n",
+       "      <td>y</td>\n",
+       "      <td>vario</td>\n",
+       "      <td>estado</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.118151</td>\n",
+       "      <td>el</td>\n",
+       "      <td>comisión</td>\n",
+       "      <td>europeo</td>\n",
+       "      <td>y</td>\n",
+       "      <td>vario</td>\n",
+       "      <td>estado</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>0.106344</td>\n",
+       "      <td>el</td>\n",
+       "      <td>comisión</td>\n",
+       "      <td>europeo</td>\n",
+       "      <td>y</td>\n",
+       "      <td>vario</td>\n",
+       "      <td>estado</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>4</td>\n",
+       "      <td>0.103996</td>\n",
+       "      <td>el</td>\n",
+       "      <td>comisión</td>\n",
+       "      <td>europeo</td>\n",
+       "      <td>y</td>\n",
+       "      <td>vario</td>\n",
+       "      <td>estado</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>5</th>\n",
+       "      <td>5</td>\n",
+       "      <td>0.118151</td>\n",
+       "      <td>el</td>\n",
+       "      <td>comisión</td>\n",
+       "      <td>europeo</td>\n",
+       "      <td>y</td>\n",
+       "      <td>vario</td>\n",
+       "      <td>estado</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>6</th>\n",
+       "      <td>6</td>\n",
+       "      <td>0.105356</td>\n",
+       "      <td>el</td>\n",
+       "      <td>comisión</td>\n",
+       "      <td>europeo</td>\n",
+       "      <td>y</td>\n",
+       "      <td>vario</td>\n",
+       "      <td>estado</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7</th>\n",
+       "      <td>7</td>\n",
+       "      <td>0.118151</td>\n",
+       "      <td>el</td>\n",
+       "      <td>comisión</td>\n",
+       "      <td>europeo</td>\n",
+       "      <td>y</td>\n",
+       "      <td>vario</td>\n",
+       "      <td>estado</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>8</th>\n",
+       "      <td>8</td>\n",
+       "      <td>0.106344</td>\n",
+       "      <td>el</td>\n",
+       "      <td>comisión</td>\n",
+       "      <td>europeo</td>\n",
+       "      <td>y</td>\n",
+       "      <td>vario</td>\n",
+       "      <td>estado</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight word1     word2    word3 word4  word5   word6\n",
+       "0     0  0.118151    el  comisión  europeo     y  vario  estado\n",
+       "1     1  0.105356    el  comisión  europeo     y  vario  estado\n",
+       "2     2  0.118151    el  comisión  europeo     y  vario  estado\n",
+       "3     3  0.106344    el  comisión  europeo     y  vario  estado\n",
+       "4     4  0.103996    el  comisión  europeo     y  vario  estado\n",
+       "5     5  0.118151    el  comisión  europeo     y  vario  estado\n",
+       "6     6  0.105356    el  comisión  europeo     y  vario  estado\n",
+       "7     7  0.118151    el  comisión  europeo     y  vario  estado\n",
+       "8     8  0.106344    el  comisión  europeo     y  vario  estado"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.029696334364774767\n",
+      "model : LinearSVM , score = 0.11456630699260187\n",
+      "model : RBFSVM , score = 0.11456630699260187\n",
+      "model : DecisionTree , score = 0.029696334364774767\n",
+      "model : RandomForest , score = 0.029696334364774767\n",
+      "model : AdaBoost , score = 0.026728736800260753\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 79+84_63+100_42+100_47_63+.csv\n",
+      "Rules(classes) number : 4\n",
+      "Words(features) number : 4\n",
+      "Records number : 252"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.296603</td>\n",
+       "      <td>no</td>\n",
+       "      <td>lo</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>hacer</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.224577</td>\n",
+       "      <td>no</td>\n",
+       "      <td>lo</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>hacer</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.249630</td>\n",
+       "      <td>no</td>\n",
+       "      <td>lo</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>hacer</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>0.229190</td>\n",
+       "      <td>no</td>\n",
+       "      <td>lo</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>hacer</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight word1 word2  word3  word4\n",
+       "0     0  0.296603    no    lo  haber  hacer\n",
+       "1     1  0.224577    no    lo  haber  hacer\n",
+       "2     2  0.249630    no    lo  haber  hacer\n",
+       "3     3  0.229190    no    lo  haber  hacer"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.21821112734051148\n",
+      "model : LinearSVM , score = 0.2624706289668316\n",
+      "model : RBFSVM , score = 0.2703155070734252\n",
+      "model : DecisionTree , score = 0.18376902613505058\n",
+      "model : RandomForest , score = 0.13061732329930537\n",
+      "model : AdaBoost , score = 0.13723675459618487\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 21+10_30+10_110_29+11_30+11_110_29+33_1_30+33_1_110_29+.csv\n",
+      "Rules(classes) number : 7\n",
+      "Words(features) number : 4\n",
+      "Records number : 2625"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.153290</td>\n",
+       "      <td>uno</td>\n",
+       "      <td>medida</td>\n",
+       "      <td>muy</td>\n",
+       "      <td>concreto</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.139238</td>\n",
+       "      <td>uno</td>\n",
+       "      <td>medida</td>\n",
+       "      <td>muy</td>\n",
+       "      <td>concreto</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.139238</td>\n",
+       "      <td>uno</td>\n",
+       "      <td>medida</td>\n",
+       "      <td>muy</td>\n",
+       "      <td>concreto</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>0.144879</td>\n",
+       "      <td>uno</td>\n",
+       "      <td>medida</td>\n",
+       "      <td>muy</td>\n",
+       "      <td>concreto</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>4</td>\n",
+       "      <td>0.144879</td>\n",
+       "      <td>uno</td>\n",
+       "      <td>medida</td>\n",
+       "      <td>muy</td>\n",
+       "      <td>concreto</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>5</th>\n",
+       "      <td>5</td>\n",
+       "      <td>0.139238</td>\n",
+       "      <td>uno</td>\n",
+       "      <td>medida</td>\n",
+       "      <td>muy</td>\n",
+       "      <td>concreto</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>6</th>\n",
+       "      <td>6</td>\n",
+       "      <td>0.139238</td>\n",
+       "      <td>uno</td>\n",
+       "      <td>medida</td>\n",
+       "      <td>muy</td>\n",
+       "      <td>concreto</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight word1   word2 word3     word4\n",
+       "0     0  0.153290   uno  medida   muy  concreto\n",
+       "1     1  0.139238   uno  medida   muy  concreto\n",
+       "2     2  0.139238   uno  medida   muy  concreto\n",
+       "3     3  0.144879   uno  medida   muy  concreto\n",
+       "4     4  0.144879   uno  medida   muy  concreto\n",
+       "5     5  0.139238   uno  medida   muy  concreto\n",
+       "6     6  0.139238   uno  medida   muy  concreto"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.13412142502798866\n",
+      "model : LinearSVM , score = 0.1353239599793647\n",
+      "model : RBFSVM , score = 0.1511010967722058\n",
+      "model : DecisionTree , score = 0.12110940368235748\n",
+      "model : RandomForest , score = 0.09958795356372663\n",
+      "model : AdaBoost , score = 0.0963204974632738\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 30+110_29+.csv\n",
+      "Rules(classes) number : 2\n",
+      "Words(features) number : 2\n",
+      "Records number : 2330"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.5</td>\n",
+       "      <td>muy</td>\n",
+       "      <td>diverso</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.5</td>\n",
+       "      <td>muy</td>\n",
+       "      <td>diverso</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule  weight word1    word2\n",
+       "0     0     0.5   muy  diverso\n",
+       "1     1     0.5   muy  diverso"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.4861290996488517\n",
+      "model : LinearSVM , score = 0.4869539176571576\n",
+      "model : RBFSVM , score = 0.3770546423512336\n",
+      "model : DecisionTree , score = 0.45048994630836775\n",
+      "model : RandomForest , score = 0.44570368285343176\n",
+      "model : AdaBoost , score = 0.4438495914597549\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 64+91+108_68+.csv\n",
+      "Rules(classes) number : 3\n",
+      "Words(features) number : 2\n",
+      "Records number : 8310"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.333705</td>\n",
+       "      <td>de</td>\n",
+       "      <td>examinar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.334584</td>\n",
+       "      <td>de</td>\n",
+       "      <td>examinar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.331712</td>\n",
+       "      <td>de</td>\n",
+       "      <td>examinar</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight word1     word2\n",
+       "0     0  0.333705    de  examinar\n",
+       "1     1  0.334584    de  examinar\n",
+       "2     2  0.331712    de  examinar"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.33265543433716654\n",
+      "model : LinearSVM , score = 0.3397881270396379\n",
+      "model : RBFSVM , score = 0.24434363205564905\n",
+      "model : DecisionTree , score = 0.3232987533102557\n",
+      "model : RandomForest , score = 0.30205680231298204\n",
+      "model : AdaBoost , score = 0.3089100161490987\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 84+100_47+.csv\n",
+      "Rules(classes) number : 2\n",
+      "Words(features) number : 3\n",
+      "Records number : 478"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.5</td>\n",
+       "      <td>también</td>\n",
+       "      <td>lo</td>\n",
+       "      <td>poner</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.5</td>\n",
+       "      <td>también</td>\n",
+       "      <td>lo</td>\n",
+       "      <td>poner</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule  weight    word1 word2  word3\n",
+       "0     0     0.5  también    lo  poner\n",
+       "1     1     0.5  también    lo  poner"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.5010828716535284\n",
+      "model : LinearSVM , score = 0.4840620915843279\n",
+      "model : RBFSVM , score = 0.34143532748351807\n",
+      "model : DecisionTree , score = 0.40254761567725456\n",
+      "model : RandomForest , score = 0.4106106872301484\n",
+      "model : AdaBoost , score = 0.4203780289178144\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 64_64+64_108_68+91_64+91_108_68+108_38+108_68_64+108_68_108_68+.csv\n",
+      "Rules(classes) number : 7\n",
+      "Words(features) number : 4\n",
+      "Records number : 14"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.141119</td>\n",
+       "      <td>de</td>\n",
+       "      <td>llegar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>cumplir</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.141119</td>\n",
+       "      <td>de</td>\n",
+       "      <td>llegar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>cumplir</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.142141</td>\n",
+       "      <td>de</td>\n",
+       "      <td>llegar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>cumplir</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>0.142141</td>\n",
+       "      <td>de</td>\n",
+       "      <td>llegar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>cumplir</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>4</td>\n",
+       "      <td>0.151996</td>\n",
+       "      <td>de</td>\n",
+       "      <td>llegar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>cumplir</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>5</th>\n",
+       "      <td>5</td>\n",
+       "      <td>0.140742</td>\n",
+       "      <td>de</td>\n",
+       "      <td>llegar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>cumplir</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>6</th>\n",
+       "      <td>6</td>\n",
+       "      <td>0.140742</td>\n",
+       "      <td>de</td>\n",
+       "      <td>llegar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>cumplir</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight word1   word2 word3    word4\n",
+       "0     0  0.141119    de  llegar     a  cumplir\n",
+       "1     1  0.141119    de  llegar     a  cumplir\n",
+       "2     2  0.142141    de  llegar     a  cumplir\n",
+       "3     3  0.142141    de  llegar     a  cumplir\n",
+       "4     4  0.151996    de  llegar     a  cumplir\n",
+       "5     5  0.140742    de  llegar     a  cumplir\n",
+       "6     6  0.140742    de  llegar     a  cumplir"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.141119\n",
+      "model : LinearSVM , score = 0.151996\n",
+      "model : RBFSVM , score = 0.151996\n",
+      "model : DecisionTree , score = 0.151996\n",
+      "model : RandomForest , score = 0.140742\n",
+      "model : AdaBoost , score = 0.151996\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 42+45_63+47_63+.csv\n",
+      "Rules(classes) number : 3\n",
+      "Words(features) number : 3\n",
+      "Records number : 21"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "/home/aboelhamd/anaconda3/lib/python3.7/site-packages/sklearn/naive_bayes.py:436: RuntimeWarning: divide by zero encountered in log\n",
+      "  n_ij = - 0.5 * np.sum(np.log(2. * np.pi * self.sigma_[i, :]))\n",
+      "/home/aboelhamd/anaconda3/lib/python3.7/site-packages/sklearn/naive_bayes.py:438: RuntimeWarning: invalid value encountered in true_divide\n",
+      "  (self.sigma_[i, :]), 1)\n"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.348117</td>\n",
+       "      <td>lo</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>aprobar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.315264</td>\n",
+       "      <td>lo</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>aprobar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.336619</td>\n",
+       "      <td>lo</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>aprobar</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight word1  word2    word3\n",
+       "0     0  0.348117    lo  haber  aprobar\n",
+       "1     1  0.315264    lo  haber  aprobar\n",
+       "2     2  0.336619    lo  haber  aprobar"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.16708374376704985\n",
+      "model : LinearSVM , score = 0.2792618308656347\n",
+      "model : RBFSVM , score = 0.2792618308656347\n",
+      "model : DecisionTree , score = 0.07951611775934861\n",
+      "model : RandomForest , score = 0.07951611775934861\n",
+      "model : AdaBoost , score = 0.07951611775934861\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 16+10_29+11_29+33_14+33_1_29+.csv\n",
+      "Rules(classes) number : 5\n",
+      "Words(features) number : 3\n",
+      "Records number : 114920"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.209393</td>\n",
+       "      <td>el</td>\n",
+       "      <td>servicio</td>\n",
+       "      <td>competente</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.194581</td>\n",
+       "      <td>el</td>\n",
+       "      <td>servicio</td>\n",
+       "      <td>competente</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.192052</td>\n",
+       "      <td>el</td>\n",
+       "      <td>servicio</td>\n",
+       "      <td>competente</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>0.209393</td>\n",
+       "      <td>el</td>\n",
+       "      <td>servicio</td>\n",
+       "      <td>competente</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>4</td>\n",
+       "      <td>0.194581</td>\n",
+       "      <td>el</td>\n",
+       "      <td>servicio</td>\n",
+       "      <td>competente</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight word1     word2       word3\n",
+       "0     0  0.209393    el  servicio  competente\n",
+       "1     1  0.194581    el  servicio  competente\n",
+       "2     2  0.192052    el  servicio  competente\n",
+       "3     3  0.209393    el  servicio  competente\n",
+       "4     4  0.194581    el  servicio  competente"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.20713068788495195\n",
+      "model : LinearSVM , score = 0.20767887864006607\n",
+      "model : RBFSVM , score = 0.19787543551643047\n",
+      "model : DecisionTree , score = 0.20371064322425628\n",
+      "model : RandomForest , score = 0.1997277746914184\n",
+      "model : AdaBoost , score = 0.20114442854828196\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 92+100_68+.csv\n",
+      "Rules(classes) number : 2\n",
+      "Words(features) number : 2\n",
+      "Records number : 380"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.505715</td>\n",
+       "      <td>no</td>\n",
+       "      <td>crear</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.494285</td>\n",
+       "      <td>no</td>\n",
+       "      <td>crear</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight word1  word2\n",
+       "0     0  0.505715    no  crear\n",
+       "1     1  0.494285    no  crear"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.4600878205646209\n",
+      "model : LinearSVM , score = 0.46085152158880804\n",
+      "model : RBFSVM , score = 0.33494255798491407\n",
+      "model : DecisionTree , score = 0.41368250856415995\n",
+      "model : RandomForest , score = 0.37487519219381\n",
+      "model : AdaBoost , score = 0.4063428873583521\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 26+16_106_29+10_29_106_29+11_29_106_29+33_24+33_14_106_29+33_1_29_106_29+.csv\n",
+      "Rules(classes) number : 7\n",
+      "Words(features) number : 5\n",
+      "Records number : 4116"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "      <th>word5</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.167642</td>\n",
+       "      <td>el</td>\n",
+       "      <td>desarrollo</td>\n",
+       "      <td>técnico</td>\n",
+       "      <td>y</td>\n",
+       "      <td>industrial</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.138759</td>\n",
+       "      <td>el</td>\n",
+       "      <td>desarrollo</td>\n",
+       "      <td>técnico</td>\n",
+       "      <td>y</td>\n",
+       "      <td>industrial</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.128489</td>\n",
+       "      <td>el</td>\n",
+       "      <td>desarrollo</td>\n",
+       "      <td>técnico</td>\n",
+       "      <td>y</td>\n",
+       "      <td>industrial</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>0.130220</td>\n",
+       "      <td>el</td>\n",
+       "      <td>desarrollo</td>\n",
+       "      <td>técnico</td>\n",
+       "      <td>y</td>\n",
+       "      <td>industrial</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>4</td>\n",
+       "      <td>0.167642</td>\n",
+       "      <td>el</td>\n",
+       "      <td>desarrollo</td>\n",
+       "      <td>técnico</td>\n",
+       "      <td>y</td>\n",
+       "      <td>industrial</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>5</th>\n",
+       "      <td>5</td>\n",
+       "      <td>0.138759</td>\n",
+       "      <td>el</td>\n",
+       "      <td>desarrollo</td>\n",
+       "      <td>técnico</td>\n",
+       "      <td>y</td>\n",
+       "      <td>industrial</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>6</th>\n",
+       "      <td>6</td>\n",
+       "      <td>0.128489</td>\n",
+       "      <td>el</td>\n",
+       "      <td>desarrollo</td>\n",
+       "      <td>técnico</td>\n",
+       "      <td>y</td>\n",
+       "      <td>industrial</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight word1       word2    word3 word4       word5\n",
+       "0     0  0.167642    el  desarrollo  técnico     y  industrial\n",
+       "1     1  0.138759    el  desarrollo  técnico     y  industrial\n",
+       "2     2  0.128489    el  desarrollo  técnico     y  industrial\n",
+       "3     3  0.130220    el  desarrollo  técnico     y  industrial\n",
+       "4     4  0.167642    el  desarrollo  técnico     y  industrial\n",
+       "5     5  0.138759    el  desarrollo  técnico     y  industrial\n",
+       "6     6  0.128489    el  desarrollo  técnico     y  industrial"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.12281064387846821\n",
+      "model : LinearSVM , score = 0.12810103192443603\n",
+      "model : RBFSVM , score = 0.14976190134123576\n",
+      "model : DecisionTree , score = 0.1090782088471264\n",
+      "model : RandomForest , score = 0.0975205407192285\n",
+      "model : AdaBoost , score = 0.10507606742700715\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 76_64+76_108_68+75_38+75_67+75_63_64+75_63_108_68+100_39+100_40_64+100_40_108_68+100_37_38+100_37_67+100_37_63_64+100_37_63_108_68+.csv\n",
+      "Rules(classes) number : 13\n",
+      "Words(features) number : 5\n",
+      "Records number : 26"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "      <th>word5</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.076446</td>\n",
+       "      <td>ni_siquiera</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>llegar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>emplear</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.076446</td>\n",
+       "      <td>ni_siquiera</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>llegar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>emplear</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.078513</td>\n",
+       "      <td>ni_siquiera</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>llegar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>emplear</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>0.076446</td>\n",
+       "      <td>ni_siquiera</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>llegar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>emplear</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>4</td>\n",
+       "      <td>0.076446</td>\n",
+       "      <td>ni_siquiera</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>llegar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>emplear</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>5</th>\n",
+       "      <td>5</td>\n",
+       "      <td>0.076446</td>\n",
+       "      <td>ni_siquiera</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>llegar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>emplear</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>6</th>\n",
+       "      <td>6</td>\n",
+       "      <td>0.078513</td>\n",
+       "      <td>ni_siquiera</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>llegar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>emplear</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7</th>\n",
+       "      <td>7</td>\n",
+       "      <td>0.076446</td>\n",
+       "      <td>ni_siquiera</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>llegar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>emplear</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>8</th>\n",
+       "      <td>8</td>\n",
+       "      <td>0.076446</td>\n",
+       "      <td>ni_siquiera</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>llegar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>emplear</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>9</th>\n",
+       "      <td>9</td>\n",
+       "      <td>0.078513</td>\n",
+       "      <td>ni_siquiera</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>llegar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>emplear</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>10</th>\n",
+       "      <td>10</td>\n",
+       "      <td>0.076446</td>\n",
+       "      <td>ni_siquiera</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>llegar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>emplear</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>11</th>\n",
+       "      <td>11</td>\n",
+       "      <td>0.076446</td>\n",
+       "      <td>ni_siquiera</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>llegar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>emplear</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>12</th>\n",
+       "      <td>12</td>\n",
+       "      <td>0.076446</td>\n",
+       "      <td>ni_siquiera</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>llegar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>emplear</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "    rule    weight        word1  word2   word3 word4    word5\n",
+       "0      0  0.076446  ni_siquiera  haber  llegar     a  emplear\n",
+       "1      1  0.076446  ni_siquiera  haber  llegar     a  emplear\n",
+       "2      2  0.078513  ni_siquiera  haber  llegar     a  emplear\n",
+       "3      3  0.076446  ni_siquiera  haber  llegar     a  emplear\n",
+       "4      4  0.076446  ni_siquiera  haber  llegar     a  emplear\n",
+       "5      5  0.076446  ni_siquiera  haber  llegar     a  emplear\n",
+       "6      6  0.078513  ni_siquiera  haber  llegar     a  emplear\n",
+       "7      7  0.076446  ni_siquiera  haber  llegar     a  emplear\n",
+       "8      8  0.076446  ni_siquiera  haber  llegar     a  emplear\n",
+       "9      9  0.078513  ni_siquiera  haber  llegar     a  emplear\n",
+       "10    10  0.076446  ni_siquiera  haber  llegar     a  emplear\n",
+       "11    11  0.076446  ni_siquiera  haber  llegar     a  emplear\n",
+       "12    12  0.076446  ni_siquiera  haber  llegar     a  emplear"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.0\n",
+      "model : LinearSVM , score = 0.07707142653346746\n",
+      "model : RBFSVM , score = 0.07707142653346746\n",
+      "model : DecisionTree , score = 0.0\n",
+      "model : RandomForest , score = 0.0\n",
+      "model : AdaBoost , score = 0.0\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 20_106+12_29_106+113_24+113_14_106+113_1_29_106+.csv\n",
+      "Rules(classes) number : 5\n",
+      "Words(features) number : 4\n",
+      "Records number : 25"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.203738</td>\n",
+       "      <td>dos</td>\n",
+       "      <td>comentario</td>\n",
+       "      <td>sencillo</td>\n",
+       "      <td>y</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.196612</td>\n",
+       "      <td>dos</td>\n",
+       "      <td>comentario</td>\n",
+       "      <td>sencillo</td>\n",
+       "      <td>y</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.199300</td>\n",
+       "      <td>dos</td>\n",
+       "      <td>comentario</td>\n",
+       "      <td>sencillo</td>\n",
+       "      <td>y</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>0.203738</td>\n",
+       "      <td>dos</td>\n",
+       "      <td>comentario</td>\n",
+       "      <td>sencillo</td>\n",
+       "      <td>y</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>4</td>\n",
+       "      <td>0.196612</td>\n",
+       "      <td>dos</td>\n",
+       "      <td>comentario</td>\n",
+       "      <td>sencillo</td>\n",
+       "      <td>y</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight word1       word2     word3 word4\n",
+       "0     0  0.203738   dos  comentario  sencillo     y\n",
+       "1     1  0.196612   dos  comentario  sencillo     y\n",
+       "2     2  0.199300   dos  comentario  sencillo     y\n",
+       "3     3  0.203738   dos  comentario  sencillo     y\n",
+       "4     4  0.196612   dos  comentario  sencillo     y"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.07628991346706404\n",
+      "model : LinearSVM , score = 0.15087718345207474\n",
+      "model : RBFSVM , score = 0.15087718345207474\n",
+      "model : DecisionTree , score = 0.0\n",
+      "model : RandomForest , score = 0.0\n",
+      "model : AdaBoost , score = 0.0\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 30_64+30_108_68+110_67+110_63_64+110_63_108_68+.csv\n",
+      "Rules(classes) number : 5\n",
+      "Words(features) number : 4\n",
+      "Records number : 35"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.204084</td>\n",
+       "      <td>bien</td>\n",
+       "      <td>informar</td>\n",
+       "      <td>para</td>\n",
+       "      <td>evitar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.193874</td>\n",
+       "      <td>bien</td>\n",
+       "      <td>informar</td>\n",
+       "      <td>para</td>\n",
+       "      <td>evitar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.204084</td>\n",
+       "      <td>bien</td>\n",
+       "      <td>informar</td>\n",
+       "      <td>para</td>\n",
+       "      <td>evitar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>0.204084</td>\n",
+       "      <td>bien</td>\n",
+       "      <td>informar</td>\n",
+       "      <td>para</td>\n",
+       "      <td>evitar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>4</td>\n",
+       "      <td>0.193874</td>\n",
+       "      <td>bien</td>\n",
+       "      <td>informar</td>\n",
+       "      <td>para</td>\n",
+       "      <td>evitar</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight word1     word2 word3   word4\n",
+       "0     0  0.204084  bien  informar  para  evitar\n",
+       "1     1  0.193874  bien  informar  para  evitar\n",
+       "2     2  0.204084  bien  informar  para  evitar\n",
+       "3     3  0.204084  bien  informar  para  evitar\n",
+       "4     4  0.193874  bien  informar  para  evitar"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.0\n",
+      "model : LinearSVM , score = 0.1692041208560311\n",
+      "model : RBFSVM , score = 0.1692041208560311\n",
+      "model : DecisionTree , score = 0.0\n",
+      "model : RandomForest , score = 0.0\n",
+      "model : AdaBoost , score = 0.0\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 80+108_92+108_100_68+.csv\n",
+      "Rules(classes) number : 3\n",
+      "Words(features) number : 3\n",
+      "Records number : 63"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.340001</td>\n",
+       "      <td>para</td>\n",
+       "      <td>no</td>\n",
+       "      <td>volver</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.334441</td>\n",
+       "      <td>para</td>\n",
+       "      <td>no</td>\n",
+       "      <td>volver</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.325558</td>\n",
+       "      <td>para</td>\n",
+       "      <td>no</td>\n",
+       "      <td>volver</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight word1 word2   word3\n",
+       "0     0  0.340001  para    no  volver\n",
+       "1     1  0.334441  para    no  volver\n",
+       "2     2  0.325558  para    no  volver"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.21988317710465202\n",
+      "model : LinearSVM , score = 0.2792714064114462\n",
+      "model : RBFSVM , score = 0.3038233460112918\n",
+      "model : DecisionTree , score = 0.15838051390647628\n",
+      "model : RandomForest , score = 0.12615808996952854\n",
+      "model : AdaBoost , score = 0.12480093199720507\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 97+108_98+108_33_99+108_33_113_108_1+.csv\n",
+      "Rules(classes) number : 4\n",
+      "Words(features) number : 5\n",
+      "Records number : 264"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "      <th>word5</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.257936</td>\n",
+       "      <td>antes_de</td>\n",
+       "      <td>el</td>\n",
+       "      <td>1</td>\n",
+       "      <td>de</td>\n",
+       "      <td>enero</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.251593</td>\n",
+       "      <td>antes_de</td>\n",
+       "      <td>el</td>\n",
+       "      <td>1</td>\n",
+       "      <td>de</td>\n",
+       "      <td>enero</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.251621</td>\n",
+       "      <td>antes_de</td>\n",
+       "      <td>el</td>\n",
+       "      <td>1</td>\n",
+       "      <td>de</td>\n",
+       "      <td>enero</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>0.238850</td>\n",
+       "      <td>antes_de</td>\n",
+       "      <td>el</td>\n",
+       "      <td>1</td>\n",
+       "      <td>de</td>\n",
+       "      <td>enero</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight     word1 word2  word3 word4  word5\n",
+       "0     0  0.257936  antes_de    el      1    de  enero\n",
+       "1     1  0.251593  antes_de    el      1    de  enero\n",
+       "2     2  0.251621  antes_de    el      1    de  enero\n",
+       "3     3  0.238850  antes_de    el      1    de  enero"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.18412234779307454\n",
+      "model : LinearSVM , score = 0.1712634434048582\n",
+      "model : RBFSVM , score = 0.2665445350564511\n",
+      "model : DecisionTree , score = 0.09966666528424266\n",
+      "model : RandomForest , score = 0.0893498872653412\n",
+      "model : AdaBoost , score = 0.12985147953020063\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 42_64+42_108_68+47_67+47_63_64+47_63_108_68+.csv\n",
+      "Rules(classes) number : 5\n",
+      "Words(features) number : 5\n",
+      "Records number : 15"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "      <th>word5</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.202631</td>\n",
+       "      <td>lo</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>hacer</td>\n",
+       "      <td>sin</td>\n",
+       "      <td>informar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.200782</td>\n",
+       "      <td>lo</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>hacer</td>\n",
+       "      <td>sin</td>\n",
+       "      <td>informar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.199458</td>\n",
+       "      <td>lo</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>hacer</td>\n",
+       "      <td>sin</td>\n",
+       "      <td>informar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>0.199458</td>\n",
+       "      <td>lo</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>hacer</td>\n",
+       "      <td>sin</td>\n",
+       "      <td>informar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>4</td>\n",
+       "      <td>0.197671</td>\n",
+       "      <td>lo</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>hacer</td>\n",
+       "      <td>sin</td>\n",
+       "      <td>informar</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight word1  word2  word3 word4     word5\n",
+       "0     0  0.202631    lo  haber  hacer   sin  informar\n",
+       "1     1  0.200782    lo  haber  hacer   sin  informar\n",
+       "2     2  0.199458    lo  haber  hacer   sin  informar\n",
+       "3     3  0.199458    lo  haber  hacer   sin  informar\n",
+       "4     4  0.197671    lo  haber  hacer   sin  informar"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.0\n",
+      "model : LinearSVM , score = 0.13698720732322098\n",
+      "model : RBFSVM , score = 0.13698720732322098\n",
+      "model : DecisionTree , score = 0.0\n",
+      "model : RandomForest , score = 0.0\n",
+      "model : AdaBoost , score = 0.0\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 30+110_63+.csv\n",
+      "Rules(classes) number : 2\n",
+      "Words(features) number : 2\n",
+      "Records number : 484"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.5</td>\n",
+       "      <td>bien</td>\n",
+       "      <td>adaptar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.5</td>\n",
+       "      <td>bien</td>\n",
+       "      <td>adaptar</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule  weight word1    word2\n",
+       "0     0     0.5  bien  adaptar\n",
+       "1     1     0.5  bien  adaptar"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.4630506070154525\n",
+      "model : LinearSVM , score = 0.4630506070154525\n",
+      "model : RBFSVM , score = 0.32572428366113676\n",
+      "model : DecisionTree , score = 0.4001060386298835\n",
+      "model : RandomForest , score = 0.38451515996862545\n",
+      "model : AdaBoost , score = 0.3845365323262198\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 98_29+33_99_29+33_113_108_14+33_113_108_1_29+.csv\n",
+      "Rules(classes) number : 4\n",
+      "Words(features) number : 5\n",
+      "Records number : 12"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "      <th>word5</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.288897</td>\n",
+       "      <td>el</td>\n",
+       "      <td>3</td>\n",
+       "      <td>de</td>\n",
+       "      <td>marzo</td>\n",
+       "      <td>pasado</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.258390</td>\n",
+       "      <td>el</td>\n",
+       "      <td>3</td>\n",
+       "      <td>de</td>\n",
+       "      <td>marzo</td>\n",
+       "      <td>pasado</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.227527</td>\n",
+       "      <td>el</td>\n",
+       "      <td>3</td>\n",
+       "      <td>de</td>\n",
+       "      <td>marzo</td>\n",
+       "      <td>pasado</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>0.225187</td>\n",
+       "      <td>el</td>\n",
+       "      <td>3</td>\n",
+       "      <td>de</td>\n",
+       "      <td>marzo</td>\n",
+       "      <td>pasado</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight word1  word2 word3  word4   word5\n",
+       "0     0  0.288897    el      3    de  marzo  pasado\n",
+       "1     1  0.258390    el      3    de  marzo  pasado\n",
+       "2     2  0.227527    el      3    de  marzo  pasado\n",
+       "3     3  0.225187    el      3    de  marzo  pasado"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.0\n",
+      "model : LinearSVM , score = 0.15216143917608507\n",
+      "model : RBFSVM , score = 0.15216143917608507\n",
+      "model : DecisionTree , score = 0.0\n",
+      "model : RandomForest , score = 0.0\n",
+      "model : AdaBoost , score = 0.0\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 35+37+.csv\n",
+      "Rules(classes) number : 2\n",
+      "Words(features) number : 1\n",
+      "Records number : 1222"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.507778</td>\n",
+       "      <td>hay</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.492222</td>\n",
+       "      <td>hay</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight word1\n",
+       "0     0  0.507778   hay\n",
+       "1     1  0.492222   hay"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.5081430489169222\n",
+      "model : LinearSVM , score = 0.5111208491983028\n",
+      "model : RBFSVM , score = 0.5111208491983028\n",
+      "model : DecisionTree , score = 0.5111208491983028\n",
+      "model : RandomForest , score = 0.5111208491983028\n",
+      "model : AdaBoost , score = 0.5111208491983028\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 17_63_29+31_15+31_1_63_29+33_29_15+33_29_1_63_29+.csv\n",
+      "Rules(classes) number : 5\n",
+      "Words(features) number : 5\n",
+      "Records number : 10"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "      <th>word5</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.201255</td>\n",
+       "      <td>el</td>\n",
+       "      <td>numeroso</td>\n",
+       "      <td>decisión</td>\n",
+       "      <td>aprobar</td>\n",
+       "      <td>relativo</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.198118</td>\n",
+       "      <td>el</td>\n",
+       "      <td>numeroso</td>\n",
+       "      <td>decisión</td>\n",
+       "      <td>aprobar</td>\n",
+       "      <td>relativo</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.201255</td>\n",
+       "      <td>el</td>\n",
+       "      <td>numeroso</td>\n",
+       "      <td>decisión</td>\n",
+       "      <td>aprobar</td>\n",
+       "      <td>relativo</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>0.198118</td>\n",
+       "      <td>el</td>\n",
+       "      <td>numeroso</td>\n",
+       "      <td>decisión</td>\n",
+       "      <td>aprobar</td>\n",
+       "      <td>relativo</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>4</td>\n",
+       "      <td>0.201255</td>\n",
+       "      <td>el</td>\n",
+       "      <td>numeroso</td>\n",
+       "      <td>decisión</td>\n",
+       "      <td>aprobar</td>\n",
+       "      <td>relativo</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight word1     word2     word3    word4     word5\n",
+       "0     0  0.201255    el  numeroso  decisión  aprobar  relativo\n",
+       "1     1  0.198118    el  numeroso  decisión  aprobar  relativo\n",
+       "2     2  0.201255    el  numeroso  decisión  aprobar  relativo\n",
+       "3     3  0.198118    el  numeroso  decisión  aprobar  relativo\n",
+       "4     4  0.201255    el  numeroso  decisión  aprobar  relativo"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.0\n",
+      "model : LinearSVM , score = 0.19984007429839853\n",
+      "model : RBFSVM , score = 0.19984007429839853\n",
+      "model : DecisionTree , score = 0.0\n",
+      "model : RandomForest , score = 0.0\n",
+      "model : AdaBoost , score = 0.0\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 20+12_29+113_14+113_1_29+.csv\n",
+      "Rules(classes) number : 4\n",
+      "Words(features) number : 3\n",
+      "Records number : 1100"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.282894</td>\n",
+       "      <td>tres</td>\n",
+       "      <td>objetivo</td>\n",
+       "      <td>principal</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.217106</td>\n",
+       "      <td>tres</td>\n",
+       "      <td>objetivo</td>\n",
+       "      <td>principal</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.282894</td>\n",
+       "      <td>tres</td>\n",
+       "      <td>objetivo</td>\n",
+       "      <td>principal</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>0.217106</td>\n",
+       "      <td>tres</td>\n",
+       "      <td>objetivo</td>\n",
+       "      <td>principal</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight word1     word2      word3\n",
+       "0     0  0.282894  tres  objetivo  principal\n",
+       "1     1  0.217106  tres  objetivo  principal\n",
+       "2     2  0.282894  tres  objetivo  principal\n",
+       "3     3  0.217106  tres  objetivo  principal"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.23407372214880567\n",
+      "model : LinearSVM , score = 0.24638366234420286\n",
+      "model : RBFSVM , score = 0.2604208440404019\n",
+      "model : DecisionTree , score = 0.21305397542331572\n",
+      "model : RandomForest , score = 0.12586922004458254\n",
+      "model : AdaBoost , score = 0.1717893312984715\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 65_64+65_108_68+64_67+64_63_64+64_63_108_68+108_40_64+108_40_108_68+108_68_67+108_68_63_64+108_68_63_108_68+.csv\n",
+      "Rules(classes) number : 10\n",
+      "Words(features) number : 5\n",
+      "Records number : 20"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "      <th>word5</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.099691</td>\n",
+       "      <td>antes_de</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>aprender</td>\n",
+       "      <td>a</td>\n",
+       "      <td>andar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.099691</td>\n",
+       "      <td>antes_de</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>aprender</td>\n",
+       "      <td>a</td>\n",
+       "      <td>andar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.099691</td>\n",
+       "      <td>antes_de</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>aprender</td>\n",
+       "      <td>a</td>\n",
+       "      <td>andar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>0.099691</td>\n",
+       "      <td>antes_de</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>aprender</td>\n",
+       "      <td>a</td>\n",
+       "      <td>andar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>4</td>\n",
+       "      <td>0.099691</td>\n",
+       "      <td>antes_de</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>aprender</td>\n",
+       "      <td>a</td>\n",
+       "      <td>andar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>5</th>\n",
+       "      <td>5</td>\n",
+       "      <td>0.100309</td>\n",
+       "      <td>antes_de</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>aprender</td>\n",
+       "      <td>a</td>\n",
+       "      <td>andar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>6</th>\n",
+       "      <td>6</td>\n",
+       "      <td>0.100309</td>\n",
+       "      <td>antes_de</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>aprender</td>\n",
+       "      <td>a</td>\n",
+       "      <td>andar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7</th>\n",
+       "      <td>7</td>\n",
+       "      <td>0.100309</td>\n",
+       "      <td>antes_de</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>aprender</td>\n",
+       "      <td>a</td>\n",
+       "      <td>andar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>8</th>\n",
+       "      <td>8</td>\n",
+       "      <td>0.100309</td>\n",
+       "      <td>antes_de</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>aprender</td>\n",
+       "      <td>a</td>\n",
+       "      <td>andar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>9</th>\n",
+       "      <td>9</td>\n",
+       "      <td>0.100309</td>\n",
+       "      <td>antes_de</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>aprender</td>\n",
+       "      <td>a</td>\n",
+       "      <td>andar</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight     word1  word2     word3 word4  word5\n",
+       "0     0  0.099691  antes_de  haber  aprender     a  andar\n",
+       "1     1  0.099691  antes_de  haber  aprender     a  andar\n",
+       "2     2  0.099691  antes_de  haber  aprender     a  andar\n",
+       "3     3  0.099691  antes_de  haber  aprender     a  andar\n",
+       "4     4  0.099691  antes_de  haber  aprender     a  andar\n",
+       "5     5  0.100309  antes_de  haber  aprender     a  andar\n",
+       "6     6  0.100309  antes_de  haber  aprender     a  andar\n",
+       "7     7  0.100309  antes_de  haber  aprender     a  andar\n",
+       "8     8  0.100309  antes_de  haber  aprender     a  andar\n",
+       "9     9  0.100309  antes_de  haber  aprender     a  andar"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.09969110030889967\n",
+      "model : LinearSVM , score = 0.10030889969110031\n",
+      "model : RBFSVM , score = 0.10030889969110031\n",
+      "model : DecisionTree , score = 0.10030889969110031\n",
+      "model : RandomForest , score = 0.09969110030889967\n",
+      "model : AdaBoost , score = 0.10030889969110031\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 74+75+100_35+100_37+.csv\n",
+      "Rules(classes) number : 4\n",
+      "Words(features) number : 2\n",
+      "Records number : 1836"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "/home/aboelhamd/anaconda3/lib/python3.7/site-packages/sklearn/naive_bayes.py:436: RuntimeWarning: divide by zero encountered in log\n",
+      "  n_ij = - 0.5 * np.sum(np.log(2. * np.pi * self.sigma_[i, :]))\n",
+      "/home/aboelhamd/anaconda3/lib/python3.7/site-packages/sklearn/naive_bayes.py:438: RuntimeWarning: invalid value encountered in true_divide\n",
+      "  (self.sigma_[i, :]), 1)\n"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.287918</td>\n",
+       "      <td>no</td>\n",
+       "      <td>hay</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.221473</td>\n",
+       "      <td>no</td>\n",
+       "      <td>hay</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.261190</td>\n",
+       "      <td>no</td>\n",
+       "      <td>hay</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>0.229420</td>\n",
+       "      <td>no</td>\n",
+       "      <td>hay</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight word1 word2\n",
+       "0     0  0.287918    no   hay\n",
+       "1     1  0.221473    no   hay\n",
+       "2     2  0.261190    no   hay\n",
+       "3     3  0.229420    no   hay"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.222433076533358\n",
+      "model : LinearSVM , score = 0.24922576304028074\n",
+      "model : RBFSVM , score = 0.23637689663456252\n",
+      "model : DecisionTree , score = 0.22220220609108216\n",
+      "model : RandomForest , score = 0.21600669587384547\n",
+      "model : AdaBoost , score = 0.22307305648417208\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 40+35_63+37_63+.csv\n",
+      "Rules(classes) number : 3\n",
+      "Words(features) number : 2\n",
+      "Records number : 924"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.333450</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>convenir</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.333099</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>convenir</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.333450</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>convenir</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight  word1     word2\n",
+       "0     0  0.333450  haber  convenir\n",
+       "1     1  0.333099  haber  convenir\n",
+       "2     2  0.333450  haber  convenir"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.33343316386434263\n",
+      "model : LinearSVM , score = 0.346860081569324\n",
+      "model : RBFSVM , score = 0.22955782483687012\n",
+      "model : DecisionTree , score = 0.29952160210106243\n",
+      "model : RandomForest , score = 0.26062750522115874\n",
+      "model : AdaBoost , score = 0.2539933509453858\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 25_29+26_14+26_1_29+16_106_29_14+16_106_29_1_29+10_29_106_29_14+10_29_106_29_1_29+11_29_106_29_14+11_29_106_29_1_29+33_23_29+33_24_14+33_24_1_29+33_14_106_29_14+33_14_106_29_1_29+33_1_29_106_29_14+33_1_29_106_29_1_29+.csv\n",
+      "Rules(classes) number : 16\n",
+      "Words(features) number : 7\n",
+      "Records number : 32"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "      <th>word5</th>\n",
+       "      <th>word6</th>\n",
+       "      <th>word7</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.063607</td>\n",
+       "      <td>el</td>\n",
+       "      <td>derecho</td>\n",
+       "      <td>humano</td>\n",
+       "      <td>y</td>\n",
+       "      <td>vario</td>\n",
+       "      <td>acontecimiento</td>\n",
+       "      <td>reciente</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.061555</td>\n",
+       "      <td>el</td>\n",
+       "      <td>derecho</td>\n",
+       "      <td>humano</td>\n",
+       "      <td>y</td>\n",
+       "      <td>vario</td>\n",
+       "      <td>acontecimiento</td>\n",
+       "      <td>reciente</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.058596</td>\n",
+       "      <td>el</td>\n",
+       "      <td>derecho</td>\n",
+       "      <td>humano</td>\n",
+       "      <td>y</td>\n",
+       "      <td>vario</td>\n",
+       "      <td>acontecimiento</td>\n",
+       "      <td>reciente</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>0.066319</td>\n",
+       "      <td>el</td>\n",
+       "      <td>derecho</td>\n",
+       "      <td>humano</td>\n",
+       "      <td>y</td>\n",
+       "      <td>vario</td>\n",
+       "      <td>acontecimiento</td>\n",
+       "      <td>reciente</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>4</td>\n",
+       "      <td>0.063607</td>\n",
+       "      <td>el</td>\n",
+       "      <td>derecho</td>\n",
+       "      <td>humano</td>\n",
+       "      <td>y</td>\n",
+       "      <td>vario</td>\n",
+       "      <td>acontecimiento</td>\n",
+       "      <td>reciente</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>5</th>\n",
+       "      <td>5</td>\n",
+       "      <td>0.063693</td>\n",
+       "      <td>el</td>\n",
+       "      <td>derecho</td>\n",
+       "      <td>humano</td>\n",
+       "      <td>y</td>\n",
+       "      <td>vario</td>\n",
+       "      <td>acontecimiento</td>\n",
+       "      <td>reciente</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>6</th>\n",
+       "      <td>6</td>\n",
+       "      <td>0.061188</td>\n",
+       "      <td>el</td>\n",
+       "      <td>derecho</td>\n",
+       "      <td>humano</td>\n",
+       "      <td>y</td>\n",
+       "      <td>vario</td>\n",
+       "      <td>acontecimiento</td>\n",
+       "      <td>reciente</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7</th>\n",
+       "      <td>7</td>\n",
+       "      <td>0.062647</td>\n",
+       "      <td>el</td>\n",
+       "      <td>derecho</td>\n",
+       "      <td>humano</td>\n",
+       "      <td>y</td>\n",
+       "      <td>vario</td>\n",
+       "      <td>acontecimiento</td>\n",
+       "      <td>reciente</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>8</th>\n",
+       "      <td>8</td>\n",
+       "      <td>0.060222</td>\n",
+       "      <td>el</td>\n",
+       "      <td>derecho</td>\n",
+       "      <td>humano</td>\n",
+       "      <td>y</td>\n",
+       "      <td>vario</td>\n",
+       "      <td>acontecimiento</td>\n",
+       "      <td>reciente</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>9</th>\n",
+       "      <td>9</td>\n",
+       "      <td>0.063607</td>\n",
+       "      <td>el</td>\n",
+       "      <td>derecho</td>\n",
+       "      <td>humano</td>\n",
+       "      <td>y</td>\n",
+       "      <td>vario</td>\n",
+       "      <td>acontecimiento</td>\n",
+       "      <td>reciente</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>10</th>\n",
+       "      <td>10</td>\n",
+       "      <td>0.061555</td>\n",
+       "      <td>el</td>\n",
+       "      <td>derecho</td>\n",
+       "      <td>humano</td>\n",
+       "      <td>y</td>\n",
+       "      <td>vario</td>\n",
+       "      <td>acontecimiento</td>\n",
+       "      <td>reciente</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>11</th>\n",
+       "      <td>11</td>\n",
+       "      <td>0.058596</td>\n",
+       "      <td>el</td>\n",
+       "      <td>derecho</td>\n",
+       "      <td>humano</td>\n",
+       "      <td>y</td>\n",
+       "      <td>vario</td>\n",
+       "      <td>acontecimiento</td>\n",
+       "      <td>reciente</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>12</th>\n",
+       "      <td>12</td>\n",
+       "      <td>0.066319</td>\n",
+       "      <td>el</td>\n",
+       "      <td>derecho</td>\n",
+       "      <td>humano</td>\n",
+       "      <td>y</td>\n",
+       "      <td>vario</td>\n",
+       "      <td>acontecimiento</td>\n",
+       "      <td>reciente</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>13</th>\n",
+       "      <td>13</td>\n",
+       "      <td>0.063607</td>\n",
+       "      <td>el</td>\n",
+       "      <td>derecho</td>\n",
+       "      <td>humano</td>\n",
+       "      <td>y</td>\n",
+       "      <td>vario</td>\n",
+       "      <td>acontecimiento</td>\n",
+       "      <td>reciente</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>14</th>\n",
+       "      <td>14</td>\n",
+       "      <td>0.063693</td>\n",
+       "      <td>el</td>\n",
+       "      <td>derecho</td>\n",
+       "      <td>humano</td>\n",
+       "      <td>y</td>\n",
+       "      <td>vario</td>\n",
+       "      <td>acontecimiento</td>\n",
+       "      <td>reciente</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>15</th>\n",
+       "      <td>15</td>\n",
+       "      <td>0.061188</td>\n",
+       "      <td>el</td>\n",
+       "      <td>derecho</td>\n",
+       "      <td>humano</td>\n",
+       "      <td>y</td>\n",
+       "      <td>vario</td>\n",
+       "      <td>acontecimiento</td>\n",
+       "      <td>reciente</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "    rule    weight word1    word2   word3 word4  word5           word6  \\\n",
+       "0      0  0.063607    el  derecho  humano     y  vario  acontecimiento   \n",
+       "1      1  0.061555    el  derecho  humano     y  vario  acontecimiento   \n",
+       "2      2  0.058596    el  derecho  humano     y  vario  acontecimiento   \n",
+       "3      3  0.066319    el  derecho  humano     y  vario  acontecimiento   \n",
+       "4      4  0.063607    el  derecho  humano     y  vario  acontecimiento   \n",
+       "5      5  0.063693    el  derecho  humano     y  vario  acontecimiento   \n",
+       "6      6  0.061188    el  derecho  humano     y  vario  acontecimiento   \n",
+       "7      7  0.062647    el  derecho  humano     y  vario  acontecimiento   \n",
+       "8      8  0.060222    el  derecho  humano     y  vario  acontecimiento   \n",
+       "9      9  0.063607    el  derecho  humano     y  vario  acontecimiento   \n",
+       "10    10  0.061555    el  derecho  humano     y  vario  acontecimiento   \n",
+       "11    11  0.058596    el  derecho  humano     y  vario  acontecimiento   \n",
+       "12    12  0.066319    el  derecho  humano     y  vario  acontecimiento   \n",
+       "13    13  0.063607    el  derecho  humano     y  vario  acontecimiento   \n",
+       "14    14  0.063693    el  derecho  humano     y  vario  acontecimiento   \n",
+       "15    15  0.061188    el  derecho  humano     y  vario  acontecimiento   \n",
+       "\n",
+       "       word7  \n",
+       "0   reciente  \n",
+       "1   reciente  \n",
+       "2   reciente  \n",
+       "3   reciente  \n",
+       "4   reciente  \n",
+       "5   reciente  \n",
+       "6   reciente  \n",
+       "7   reciente  \n",
+       "8   reciente  \n",
+       "9   reciente  \n",
+       "10  reciente  \n",
+       "11  reciente  \n",
+       "12  reciente  \n",
+       "13  reciente  \n",
+       "14  reciente  \n",
+       "15  reciente  "
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.0\n",
+      "model : LinearSVM , score = 0.06659291686619512\n",
+      "model : RBFSVM , score = 0.06659291686619512\n",
+      "model : DecisionTree , score = 0.0\n",
+      "model : RandomForest , score = 0.0\n",
+      "model : AdaBoost , score = 0.0\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 76+74_63+75_63+100_40+100_35_63+100_37_63+.csv\n",
+      "Rules(classes) number : 6\n",
+      "Words(features) number : 3\n",
+      "Records number : 618"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.17124</td>\n",
+       "      <td>quizá</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>llegar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.15752</td>\n",
+       "      <td>quizá</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>llegar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.17124</td>\n",
+       "      <td>quizá</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>llegar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>0.17124</td>\n",
+       "      <td>quizá</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>llegar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>4</td>\n",
+       "      <td>0.15752</td>\n",
+       "      <td>quizá</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>llegar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>5</th>\n",
+       "      <td>5</td>\n",
+       "      <td>0.17124</td>\n",
+       "      <td>quizá</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>llegar</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule   weight  word1  word2   word3\n",
+       "0     0  0.17124  quizá  haber  llegar\n",
+       "1     1  0.15752  quizá  haber  llegar\n",
+       "2     2  0.17124  quizá  haber  llegar\n",
+       "3     3  0.17124  quizá  haber  llegar\n",
+       "4     4  0.15752  quizá  haber  llegar\n",
+       "5     5  0.17124  quizá  haber  llegar"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.15541247961251903\n",
+      "model : LinearSVM , score = 0.15642422260752123\n",
+      "model : RBFSVM , score = 0.1781171815647689\n",
+      "model : DecisionTree , score = 0.11920340132556226\n",
+      "model : RandomForest , score = 0.07302332395743687\n",
+      "model : AdaBoost , score = 0.08910207760680809\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 19+16_29+10_29_29+11_29_29+33_15+33_14_29+33_1_29_29+.csv\n",
+      "Rules(classes) number : 7\n",
+      "Words(features) number : 4\n",
+      "Records number : 6734"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.146545</td>\n",
+       "      <td>ninguno</td>\n",
+       "      <td>obstáculo</td>\n",
+       "      <td>jurídico</td>\n",
+       "      <td>real</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.140010</td>\n",
+       "      <td>ninguno</td>\n",
+       "      <td>obstáculo</td>\n",
+       "      <td>jurídico</td>\n",
+       "      <td>real</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.143535</td>\n",
+       "      <td>ninguno</td>\n",
+       "      <td>obstáculo</td>\n",
+       "      <td>jurídico</td>\n",
+       "      <td>real</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>0.139819</td>\n",
+       "      <td>ninguno</td>\n",
+       "      <td>obstáculo</td>\n",
+       "      <td>jurídico</td>\n",
+       "      <td>real</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>4</td>\n",
+       "      <td>0.146545</td>\n",
+       "      <td>ninguno</td>\n",
+       "      <td>obstáculo</td>\n",
+       "      <td>jurídico</td>\n",
+       "      <td>real</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>5</th>\n",
+       "      <td>5</td>\n",
+       "      <td>0.140010</td>\n",
+       "      <td>ninguno</td>\n",
+       "      <td>obstáculo</td>\n",
+       "      <td>jurídico</td>\n",
+       "      <td>real</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>6</th>\n",
+       "      <td>6</td>\n",
+       "      <td>0.143535</td>\n",
+       "      <td>ninguno</td>\n",
+       "      <td>obstáculo</td>\n",
+       "      <td>jurídico</td>\n",
+       "      <td>real</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight    word1      word2     word3 word4\n",
+       "0     0  0.146545  ninguno  obstáculo  jurídico  real\n",
+       "1     1  0.140010  ninguno  obstáculo  jurídico  real\n",
+       "2     2  0.143535  ninguno  obstáculo  jurídico  real\n",
+       "3     3  0.139819  ninguno  obstáculo  jurídico  real\n",
+       "4     4  0.146545  ninguno  obstáculo  jurídico  real\n",
+       "5     5  0.140010  ninguno  obstáculo  jurídico  real\n",
+       "6     6  0.143535  ninguno  obstáculo  jurídico  real"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.13357139094346357\n",
+      "model : LinearSVM , score = 0.14457829516012763\n",
+      "model : RBFSVM , score = 0.14829315087623948\n",
+      "model : DecisionTree , score = 0.1286554919938203\n",
+      "model : RandomForest , score = 0.09791982020184645\n",
+      "model : AdaBoost , score = 0.10583255654691229\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 93+100_69+.csv\n",
+      "Rules(classes) number : 2\n",
+      "Words(features) number : 2\n",
+      "Records number : 14"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.503313</td>\n",
+       "      <td>no</td>\n",
+       "      <td>hacer</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.496687</td>\n",
+       "      <td>no</td>\n",
+       "      <td>hacer</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight word1  word2\n",
+       "0     0  0.503313    no  hacer\n",
+       "1     1  0.496687    no  hacer"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.28706249003625595\n",
+      "model : LinearSVM , score = 0.4242505909801415\n",
+      "model : RBFSVM , score = 0.4242505909801415\n",
+      "model : DecisionTree , score = 0.2855692474154155\n",
+      "model : RandomForest , score = 0.28331284259950834\n",
+      "model : AdaBoost , score = 0.2855692474154155\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 28+16_106_30+16_106_100_29+10_29_106_30+10_29_106_100_29+11_29_106_30+11_29_106_100_29+33_14_106_30+33_14_106_100_29+33_1_29_106_30+33_1_29_106_100_29+.csv\n",
+      "Rules(classes) number : 11\n",
+      "Words(features) number : 6\n",
+      "Records number : 242"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "      <th>word5</th>\n",
+       "      <th>word6</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.091505</td>\n",
+       "      <td>uno</td>\n",
+       "      <td>debate</td>\n",
+       "      <td>abierto</td>\n",
+       "      <td>y</td>\n",
+       "      <td>mucho_más</td>\n",
+       "      <td>dinámico</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.092972</td>\n",
+       "      <td>uno</td>\n",
+       "      <td>debate</td>\n",
+       "      <td>abierto</td>\n",
+       "      <td>y</td>\n",
+       "      <td>mucho_más</td>\n",
+       "      <td>dinámico</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.092972</td>\n",
+       "      <td>uno</td>\n",
+       "      <td>debate</td>\n",
+       "      <td>abierto</td>\n",
+       "      <td>y</td>\n",
+       "      <td>mucho_más</td>\n",
+       "      <td>dinámico</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>0.089077</td>\n",
+       "      <td>uno</td>\n",
+       "      <td>debate</td>\n",
+       "      <td>abierto</td>\n",
+       "      <td>y</td>\n",
+       "      <td>mucho_más</td>\n",
+       "      <td>dinámico</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>4</td>\n",
+       "      <td>0.089077</td>\n",
+       "      <td>uno</td>\n",
+       "      <td>debate</td>\n",
+       "      <td>abierto</td>\n",
+       "      <td>y</td>\n",
+       "      <td>mucho_más</td>\n",
+       "      <td>dinámico</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>5</th>\n",
+       "      <td>5</td>\n",
+       "      <td>0.090149</td>\n",
+       "      <td>uno</td>\n",
+       "      <td>debate</td>\n",
+       "      <td>abierto</td>\n",
+       "      <td>y</td>\n",
+       "      <td>mucho_más</td>\n",
+       "      <td>dinámico</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>6</th>\n",
+       "      <td>6</td>\n",
+       "      <td>0.090149</td>\n",
+       "      <td>uno</td>\n",
+       "      <td>debate</td>\n",
+       "      <td>abierto</td>\n",
+       "      <td>y</td>\n",
+       "      <td>mucho_más</td>\n",
+       "      <td>dinámico</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7</th>\n",
+       "      <td>7</td>\n",
+       "      <td>0.092972</td>\n",
+       "      <td>uno</td>\n",
+       "      <td>debate</td>\n",
+       "      <td>abierto</td>\n",
+       "      <td>y</td>\n",
+       "      <td>mucho_más</td>\n",
+       "      <td>dinámico</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>8</th>\n",
+       "      <td>8</td>\n",
+       "      <td>0.092972</td>\n",
+       "      <td>uno</td>\n",
+       "      <td>debate</td>\n",
+       "      <td>abierto</td>\n",
+       "      <td>y</td>\n",
+       "      <td>mucho_más</td>\n",
+       "      <td>dinámico</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>9</th>\n",
+       "      <td>9</td>\n",
+       "      <td>0.089077</td>\n",
+       "      <td>uno</td>\n",
+       "      <td>debate</td>\n",
+       "      <td>abierto</td>\n",
+       "      <td>y</td>\n",
+       "      <td>mucho_más</td>\n",
+       "      <td>dinámico</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>10</th>\n",
+       "      <td>10</td>\n",
+       "      <td>0.089077</td>\n",
+       "      <td>uno</td>\n",
+       "      <td>debate</td>\n",
+       "      <td>abierto</td>\n",
+       "      <td>y</td>\n",
+       "      <td>mucho_más</td>\n",
+       "      <td>dinámico</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "    rule    weight word1   word2    word3 word4      word5     word6\n",
+       "0      0  0.091505   uno  debate  abierto     y  mucho_más  dinámico\n",
+       "1      1  0.092972   uno  debate  abierto     y  mucho_más  dinámico\n",
+       "2      2  0.092972   uno  debate  abierto     y  mucho_más  dinámico\n",
+       "3      3  0.089077   uno  debate  abierto     y  mucho_más  dinámico\n",
+       "4      4  0.089077   uno  debate  abierto     y  mucho_más  dinámico\n",
+       "5      5  0.090149   uno  debate  abierto     y  mucho_más  dinámico\n",
+       "6      6  0.090149   uno  debate  abierto     y  mucho_más  dinámico\n",
+       "7      7  0.092972   uno  debate  abierto     y  mucho_más  dinámico\n",
+       "8      8  0.092972   uno  debate  abierto     y  mucho_más  dinámico\n",
+       "9      9  0.089077   uno  debate  abierto     y  mucho_más  dinámico\n",
+       "10    10  0.089077   uno  debate  abierto     y  mucho_más  dinámico"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.02499643012919274\n",
+      "model : LinearSVM , score = 0.05853263185872462\n",
+      "model : RBFSVM , score = 0.09487956768291234\n",
+      "model : DecisionTree , score = 0.0\n",
+      "model : RandomForest , score = 0.0\n",
+      "model : AdaBoost , score = 0.04976535493069682\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 40+63_63+.csv\n",
+      "Rules(classes) number : 2\n",
+      "Words(features) number : 2\n",
+      "Records number : 4"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.486204</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>repetir</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.513796</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>repetir</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight  word1    word2\n",
+       "0     0  0.486204  haber  repetir\n",
+       "1     1  0.513796  haber  repetir"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.48620399999999997\n",
+      "model : LinearSVM , score = 0.513796\n",
+      "model : RBFSVM , score = 0.513796\n",
+      "model : DecisionTree , score = 0.513796\n",
+      "model : RandomForest , score = 0.513796\n",
+      "model : AdaBoost , score = 0.513796\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 98+33_99+33_113_108_1+.csv\n",
+      "Rules(classes) number : 3\n",
+      "Words(features) number : 4\n",
+      "Records number : 213"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "/home/aboelhamd/anaconda3/lib/python3.7/site-packages/sklearn/naive_bayes.py:436: RuntimeWarning: divide by zero encountered in log\n",
+      "  n_ij = - 0.5 * np.sum(np.log(2. * np.pi * self.sigma_[i, :]))\n",
+      "/home/aboelhamd/anaconda3/lib/python3.7/site-packages/sklearn/naive_bayes.py:438: RuntimeWarning: invalid value encountered in true_divide\n",
+      "  (self.sigma_[i, :]), 1)\n"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.336429</td>\n",
+       "      <td>el</td>\n",
+       "      <td>27</td>\n",
+       "      <td>de</td>\n",
+       "      <td>marzo</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.338076</td>\n",
+       "      <td>el</td>\n",
+       "      <td>27</td>\n",
+       "      <td>de</td>\n",
+       "      <td>marzo</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.325496</td>\n",
+       "      <td>el</td>\n",
+       "      <td>27</td>\n",
+       "      <td>de</td>\n",
+       "      <td>marzo</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight word1 word2 word3  word4\n",
+       "0     0  0.336429    el    27    de  marzo\n",
+       "1     1  0.338076    el    27    de  marzo\n",
+       "2     2  0.325496    el    27    de  marzo"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.2851337068878107\n",
+      "model : LinearSVM , score = 0.26689440142665044\n",
+      "model : RBFSVM , score = 0.32354741596092174\n",
+      "model : DecisionTree , score = 0.130059015477188\n",
+      "model : RandomForest , score = 0.13016797136447059\n",
+      "model : AdaBoost , score = 0.17396683289488685\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 40+37_63+.csv\n",
+      "Rules(classes) number : 2\n",
+      "Words(features) number : 2\n",
+      "Records number : 11700"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.5</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>incluir</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.5</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>incluir</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule  weight  word1    word2\n",
+       "0     0     0.5  haber  incluir\n",
+       "1     1     0.5  haber  incluir"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.4917035327565232\n",
+      "model : LinearSVM , score = 0.4933275214113945\n",
+      "model : RBFSVM , score = 0.43242836955937985\n",
+      "model : DecisionTree , score = 0.4761490467706863\n",
+      "model : RandomForest , score = 0.47036086996205184\n",
+      "model : AdaBoost , score = 0.46738221547011866\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 75+100_37+.csv\n",
+      "Rules(classes) number : 2\n",
+      "Words(features) number : 2\n",
+      "Records number : 23248"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.515855</td>\n",
+       "      <td>no</td>\n",
+       "      <td>respetar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.484145</td>\n",
+       "      <td>no</td>\n",
+       "      <td>respetar</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight word1     word2\n",
+       "0     0  0.515855    no  respetar\n",
+       "1     1  0.484145    no  respetar"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.5018283420998348\n",
+      "model : LinearSVM , score = 0.5043203550155587\n",
+      "model : RBFSVM , score = 0.41599039176105\n",
+      "model : DecisionTree , score = 0.49078794172529744\n",
+      "model : RandomForest , score = 0.4824725030459211\n",
+      "model : AdaBoost , score = 0.48468805774189694\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 20+12_29+113_15+113_14+113_1_29+.csv\n",
+      "Rules(classes) number : 5\n",
+      "Words(features) number : 3\n",
+      "Records number : 45"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.200115</td>\n",
+       "      <td>20.000</td>\n",
+       "      <td>acto</td>\n",
+       "      <td>jurídico</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.200115</td>\n",
+       "      <td>20.000</td>\n",
+       "      <td>acto</td>\n",
+       "      <td>jurídico</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.199540</td>\n",
+       "      <td>20.000</td>\n",
+       "      <td>acto</td>\n",
+       "      <td>jurídico</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>0.200115</td>\n",
+       "      <td>20.000</td>\n",
+       "      <td>acto</td>\n",
+       "      <td>jurídico</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>4</td>\n",
+       "      <td>0.200115</td>\n",
+       "      <td>20.000</td>\n",
+       "      <td>acto</td>\n",
+       "      <td>jurídico</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight   word1 word2     word3\n",
+       "0     0  0.200115  20.000  acto  jurídico\n",
+       "1     1  0.200115  20.000  acto  jurídico\n",
+       "2     2  0.199540  20.000  acto  jurídico\n",
+       "3     3  0.200115  20.000  acto  jurídico\n",
+       "4     4  0.200115  20.000  acto  jurídico"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.1309606202361131\n",
+      "model : LinearSVM , score = 0.1751275346500531\n",
+      "model : RBFSVM , score = 0.1751275346500531\n",
+      "model : DecisionTree , score = 0.04407809457336834\n",
+      "model : RandomForest , score = 0.04407809457336834\n",
+      "model : AdaBoost , score = 0.04407809457336834\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 17+31_1+33_29_1+.csv\n",
+      "Rules(classes) number : 3\n",
+      "Words(features) number : 3\n",
+      "Records number : 20829"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.333333</td>\n",
+       "      <td>uno</td>\n",
+       "      <td>bueno</td>\n",
+       "      <td>fórmula</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.333333</td>\n",
+       "      <td>uno</td>\n",
+       "      <td>bueno</td>\n",
+       "      <td>fórmula</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.333333</td>\n",
+       "      <td>uno</td>\n",
+       "      <td>bueno</td>\n",
+       "      <td>fórmula</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight word1  word2    word3\n",
+       "0     0  0.333333   uno  bueno  fórmula\n",
+       "1     1  0.333333   uno  bueno  fórmula\n",
+       "2     2  0.333333   uno  bueno  fórmula"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.3295921616760996\n",
+      "model : LinearSVM , score = 0.32920708495257434\n",
+      "model : RBFSVM , score = 0.14985848799535995\n",
+      "model : DecisionTree , score = 0.3181969882034842\n",
+      "model : RandomForest , score = 0.3066639449713755\n",
+      "model : AdaBoost , score = 0.30670845232364624\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 32+33_30+33_100_63+.csv\n",
+      "Rules(classes) number : 3\n",
+      "Words(features) number : 3\n",
+      "Records number : 51"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.336110</td>\n",
+       "      <td>el</td>\n",
+       "      <td>más</td>\n",
+       "      <td>preocupar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.331945</td>\n",
+       "      <td>el</td>\n",
+       "      <td>más</td>\n",
+       "      <td>preocupar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.331945</td>\n",
+       "      <td>el</td>\n",
+       "      <td>más</td>\n",
+       "      <td>preocupar</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight word1 word2      word3\n",
+       "0     0  0.336110    el   más  preocupar\n",
+       "1     1  0.331945    el   más  preocupar\n",
+       "2     2  0.331945    el   más  preocupar"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.2300763669196149\n",
+      "model : LinearSVM , score = 0.30457753516806924\n",
+      "model : RBFSVM , score = 0.30457753516806924\n",
+      "model : DecisionTree , score = 0.15145278645061538\n",
+      "model : RandomForest , score = 0.1906317271927346\n",
+      "model : AdaBoost , score = 0.18972779925884573\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 76_64+76_108_68+74_67+74_63_64+74_63_108_68+75_67+75_63_64+75_63_108_68+100_40_64+100_40_108_68+100_35_67+100_35_63_64+100_35_63_108_68+100_37_67+100_37_63_64+100_37_63_108_68+.csv\n",
+      "Rules(classes) number : 16\n",
+      "Words(features) number : 5\n",
+      "Records number : 32"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "      <th>word5</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.066377</td>\n",
+       "      <td>ni_siquiera</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>empezar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>trabajar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.066377</td>\n",
+       "      <td>ni_siquiera</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>empezar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>trabajar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.056037</td>\n",
+       "      <td>ni_siquiera</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>empezar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>trabajar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>0.056037</td>\n",
+       "      <td>ni_siquiera</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>empezar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>trabajar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>4</td>\n",
+       "      <td>0.056037</td>\n",
+       "      <td>ni_siquiera</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>empezar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>trabajar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>5</th>\n",
+       "      <td>5</td>\n",
+       "      <td>0.066377</td>\n",
+       "      <td>ni_siquiera</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>empezar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>trabajar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>6</th>\n",
+       "      <td>6</td>\n",
+       "      <td>0.066377</td>\n",
+       "      <td>ni_siquiera</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>empezar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>trabajar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7</th>\n",
+       "      <td>7</td>\n",
+       "      <td>0.066377</td>\n",
+       "      <td>ni_siquiera</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>empezar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>trabajar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>8</th>\n",
+       "      <td>8</td>\n",
+       "      <td>0.066377</td>\n",
+       "      <td>ni_siquiera</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>empezar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>trabajar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>9</th>\n",
+       "      <td>9</td>\n",
+       "      <td>0.066377</td>\n",
+       "      <td>ni_siquiera</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>empezar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>trabajar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>10</th>\n",
+       "      <td>10</td>\n",
+       "      <td>0.056037</td>\n",
+       "      <td>ni_siquiera</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>empezar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>trabajar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>11</th>\n",
+       "      <td>11</td>\n",
+       "      <td>0.056037</td>\n",
+       "      <td>ni_siquiera</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>empezar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>trabajar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>12</th>\n",
+       "      <td>12</td>\n",
+       "      <td>0.056037</td>\n",
+       "      <td>ni_siquiera</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>empezar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>trabajar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>13</th>\n",
+       "      <td>13</td>\n",
+       "      <td>0.066377</td>\n",
+       "      <td>ni_siquiera</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>empezar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>trabajar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>14</th>\n",
+       "      <td>14</td>\n",
+       "      <td>0.066377</td>\n",
+       "      <td>ni_siquiera</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>empezar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>trabajar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>15</th>\n",
+       "      <td>15</td>\n",
+       "      <td>0.066377</td>\n",
+       "      <td>ni_siquiera</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>empezar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>trabajar</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "    rule    weight        word1  word2    word3 word4     word5\n",
+       "0      0  0.066377  ni_siquiera  haber  empezar     a  trabajar\n",
+       "1      1  0.066377  ni_siquiera  haber  empezar     a  trabajar\n",
+       "2      2  0.056037  ni_siquiera  haber  empezar     a  trabajar\n",
+       "3      3  0.056037  ni_siquiera  haber  empezar     a  trabajar\n",
+       "4      4  0.056037  ni_siquiera  haber  empezar     a  trabajar\n",
+       "5      5  0.066377  ni_siquiera  haber  empezar     a  trabajar\n",
+       "6      6  0.066377  ni_siquiera  haber  empezar     a  trabajar\n",
+       "7      7  0.066377  ni_siquiera  haber  empezar     a  trabajar\n",
+       "8      8  0.066377  ni_siquiera  haber  empezar     a  trabajar\n",
+       "9      9  0.066377  ni_siquiera  haber  empezar     a  trabajar\n",
+       "10    10  0.056037  ni_siquiera  haber  empezar     a  trabajar\n",
+       "11    11  0.056037  ni_siquiera  haber  empezar     a  trabajar\n",
+       "12    12  0.056037  ni_siquiera  haber  empezar     a  trabajar\n",
+       "13    13  0.066377  ni_siquiera  haber  empezar     a  trabajar\n",
+       "14    14  0.066377  ni_siquiera  haber  empezar     a  trabajar\n",
+       "15    15  0.066377  ni_siquiera  haber  empezar     a  trabajar"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.06637753982652389\n",
+      "model : LinearSVM , score = 0.06637753982652389\n",
+      "model : RBFSVM , score = 0.06637753982652389\n",
+      "model : DecisionTree , score = 0.06637753982652389\n",
+      "model : RandomForest , score = 0.06637753982652389\n",
+      "model : AdaBoost , score = 0.06637753982652389\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 30_64+30_108_68+100_67+100_63_64+100_63_108_68+.csv\n",
+      "Rules(classes) number : 5\n",
+      "Words(features) number : 4\n",
+      "Records number : 105"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "/home/aboelhamd/anaconda3/lib/python3.7/site-packages/sklearn/naive_bayes.py:436: RuntimeWarning: divide by zero encountered in log\n",
+      "  n_ij = - 0.5 * np.sum(np.log(2. * np.pi * self.sigma_[i, :]))\n",
+      "/home/aboelhamd/anaconda3/lib/python3.7/site-packages/sklearn/naive_bayes.py:438: RuntimeWarning: invalid value encountered in true_divide\n",
+      "  (self.sigma_[i, :]), 1)\n"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.2</td>\n",
+       "      <td>realmente</td>\n",
+       "      <td>disponer</td>\n",
+       "      <td>a</td>\n",
+       "      <td>librar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.2</td>\n",
+       "      <td>realmente</td>\n",
+       "      <td>disponer</td>\n",
+       "      <td>a</td>\n",
+       "      <td>librar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.2</td>\n",
+       "      <td>realmente</td>\n",
+       "      <td>disponer</td>\n",
+       "      <td>a</td>\n",
+       "      <td>librar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>0.2</td>\n",
+       "      <td>realmente</td>\n",
+       "      <td>disponer</td>\n",
+       "      <td>a</td>\n",
+       "      <td>librar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>4</td>\n",
+       "      <td>0.2</td>\n",
+       "      <td>realmente</td>\n",
+       "      <td>disponer</td>\n",
+       "      <td>a</td>\n",
+       "      <td>librar</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule  weight      word1     word2 word3   word4\n",
+       "0     0     0.2  realmente  disponer     a  librar\n",
+       "1     1     0.2  realmente  disponer     a  librar\n",
+       "2     2     0.2  realmente  disponer     a  librar\n",
+       "3     3     0.2  realmente  disponer     a  librar\n",
+       "4     4     0.2  realmente  disponer     a  librar"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.1131868758808217\n",
+      "model : LinearSVM , score = 0.18936578010447513\n",
+      "model : RBFSVM , score = 0.18913502781145083\n",
+      "model : DecisionTree , score = 0.05658834573615122\n",
+      "model : RandomForest , score = 0.018868314183344823\n",
+      "model : AdaBoost , score = 0.07544836151255443\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 64+108_68+.csv\n",
+      "Rules(classes) number : 2\n",
+      "Words(features) number : 2\n",
+      "Records number : 10714"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.516806</td>\n",
+       "      <td>por</td>\n",
+       "      <td>considerar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.483194</td>\n",
+       "      <td>por</td>\n",
+       "      <td>considerar</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight word1       word2\n",
+       "0     0  0.516806   por  considerar\n",
+       "1     1  0.483194   por  considerar"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.4929621418967288\n",
+      "model : LinearSVM , score = 0.4942816164223336\n",
+      "model : RBFSVM , score = 0.37833582985460873\n",
+      "model : DecisionTree , score = 0.4755386350131668\n",
+      "model : RandomForest , score = 0.47136806950311533\n",
+      "model : AdaBoost , score = 0.4797982398098202\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 45+47+.csv\n",
+      "Rules(classes) number : 2\n",
+      "Words(features) number : 2\n",
+      "Records number : 6"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.467616</td>\n",
+       "      <td>lo</td>\n",
+       "      <td>haber</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.532384</td>\n",
+       "      <td>lo</td>\n",
+       "      <td>haber</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight word1  word2\n",
+       "0     0  0.467616    lo  haber\n",
+       "1     1  0.532384    lo  haber"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.6510258814294746\n",
+      "model : LinearSVM , score = 0.3489741185705253\n",
+      "model : RBFSVM , score = 0.3489741185705253\n",
+      "model : DecisionTree , score = 0.3489741185705253\n",
+      "model : RandomForest , score = 0.3489741185705253\n",
+      "model : AdaBoost , score = 0.3489741185705253\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 38+69_64+69_108_68+.csv\n",
+      "Rules(classes) number : 3\n",
+      "Words(features) number : 3\n",
+      "Records number : 6"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "/home/aboelhamd/anaconda3/lib/python3.7/site-packages/sklearn/naive_bayes.py:436: RuntimeWarning: divide by zero encountered in log\n",
+      "  n_ij = - 0.5 * np.sum(np.log(2. * np.pi * self.sigma_[i, :]))\n",
+      "/home/aboelhamd/anaconda3/lib/python3.7/site-packages/sklearn/naive_bayes.py:438: RuntimeWarning: divide by zero encountered in true_divide\n",
+      "  (self.sigma_[i, :]), 1)\n",
+      "/home/aboelhamd/anaconda3/lib/python3.7/site-packages/sklearn/naive_bayes.py:438: RuntimeWarning: invalid value encountered in true_divide\n",
+      "  (self.sigma_[i, :]), 1)\n"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.349408</td>\n",
+       "      <td>llegar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>ser</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.325296</td>\n",
+       "      <td>llegar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>ser</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.325296</td>\n",
+       "      <td>llegar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>ser</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight   word1 word2 word3\n",
+       "0     0  0.349408  llegar     a   ser\n",
+       "1     1  0.325296  llegar     a   ser\n",
+       "2     2  0.325296  llegar     a   ser"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.349408\n",
+      "model : LinearSVM , score = 0.349408\n",
+      "model : RBFSVM , score = 0.349408\n",
+      "model : DecisionTree , score = 0.349408\n",
+      "model : RandomForest , score = 0.349408\n",
+      "model : AdaBoost , score = 0.349408\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 65+64_63+91_63+108_40+108_68_63+.csv\n",
+      "Rules(classes) number : 5\n",
+      "Words(features) number : 3\n",
+      "Records number : 90"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "/home/aboelhamd/anaconda3/lib/python3.7/site-packages/sklearn/naive_bayes.py:436: RuntimeWarning: divide by zero encountered in log\n",
+      "  n_ij = - 0.5 * np.sum(np.log(2. * np.pi * self.sigma_[i, :]))\n",
+      "/home/aboelhamd/anaconda3/lib/python3.7/site-packages/sklearn/naive_bayes.py:438: RuntimeWarning: invalid value encountered in true_divide\n",
+      "  (self.sigma_[i, :]), 1)\n"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.198241</td>\n",
+       "      <td>de</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>ser</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.201699</td>\n",
+       "      <td>de</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>ser</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.202473</td>\n",
+       "      <td>de</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>ser</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>0.198794</td>\n",
+       "      <td>de</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>ser</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>4</td>\n",
+       "      <td>0.198794</td>\n",
+       "      <td>de</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>ser</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight word1  word2 word3\n",
+       "0     0  0.198241    de  haber   ser\n",
+       "1     1  0.201699    de  haber   ser\n",
+       "2     2  0.202473    de  haber   ser\n",
+       "3     3  0.198794    de  haber   ser\n",
+       "4     4  0.198794    de  haber   ser"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.15721598781487842\n",
+      "model : LinearSVM , score = 0.13246373705958808\n",
+      "model : RBFSVM , score = 0.20180785092965567\n",
+      "model : DecisionTree , score = 0.04382684750214566\n",
+      "model : RandomForest , score = 0.021909366934298014\n",
+      "model : AdaBoost , score = 0.08857531485899726\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 40_64+40_108_68+35_67+35_63_64+35_63_108_68+37_67+37_63_64+37_63_108_68+.csv\n",
+      "Rules(classes) number : 8\n",
+      "Words(features) number : 4\n",
+      "Records number : 16"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.133838</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>amenazar</td>\n",
+       "      <td>con</td>\n",
+       "      <td>recurrir</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.130860</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>amenazar</td>\n",
+       "      <td>con</td>\n",
+       "      <td>recurrir</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.112964</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>amenazar</td>\n",
+       "      <td>con</td>\n",
+       "      <td>recurrir</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>0.112964</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>amenazar</td>\n",
+       "      <td>con</td>\n",
+       "      <td>recurrir</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>4</td>\n",
+       "      <td>0.110838</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>amenazar</td>\n",
+       "      <td>con</td>\n",
+       "      <td>recurrir</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>5</th>\n",
+       "      <td>5</td>\n",
+       "      <td>0.133838</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>amenazar</td>\n",
+       "      <td>con</td>\n",
+       "      <td>recurrir</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>6</th>\n",
+       "      <td>6</td>\n",
+       "      <td>0.133838</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>amenazar</td>\n",
+       "      <td>con</td>\n",
+       "      <td>recurrir</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7</th>\n",
+       "      <td>7</td>\n",
+       "      <td>0.130860</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>amenazar</td>\n",
+       "      <td>con</td>\n",
+       "      <td>recurrir</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight  word1     word2 word3     word4\n",
+       "0     0  0.133838  haber  amenazar   con  recurrir\n",
+       "1     1  0.130860  haber  amenazar   con  recurrir\n",
+       "2     2  0.112964  haber  amenazar   con  recurrir\n",
+       "3     3  0.112964  haber  amenazar   con  recurrir\n",
+       "4     4  0.110838  haber  amenazar   con  recurrir\n",
+       "5     5  0.133838  haber  amenazar   con  recurrir\n",
+       "6     6  0.133838  haber  amenazar   con  recurrir\n",
+       "7     7  0.130860  haber  amenazar   con  recurrir"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.0\n",
+      "model : LinearSVM , score = 0.1264077214410737\n",
+      "model : RBFSVM , score = 0.1264077214410737\n",
+      "model : DecisionTree , score = 0.0\n",
+      "model : RandomForest , score = 0.0\n",
+      "model : AdaBoost , score = 0.0\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 7_63_106_29+8_63_106_29+9_63_106_29+10_108_24+10_108_1_63_106_29+11_108_24+11_108_1_63_106_29+33_1_108_24+33_1_108_1_63_106_29+.csv\n",
+      "Rules(classes) number : 9\n",
+      "Words(features) number : 7\n",
+      "Records number : 45"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "      <th>word5</th>\n",
+       "      <th>word6</th>\n",
+       "      <th>word7</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.111574</td>\n",
+       "      <td>uno</td>\n",
+       "      <td>definición</td>\n",
+       "      <td>de</td>\n",
+       "      <td>ciudadano</td>\n",
+       "      <td>ampliar</td>\n",
+       "      <td>y</td>\n",
+       "      <td>confuso</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.110444</td>\n",
+       "      <td>uno</td>\n",
+       "      <td>definición</td>\n",
+       "      <td>de</td>\n",
+       "      <td>ciudadano</td>\n",
+       "      <td>ampliar</td>\n",
+       "      <td>y</td>\n",
+       "      <td>confuso</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.111013</td>\n",
+       "      <td>uno</td>\n",
+       "      <td>definición</td>\n",
+       "      <td>de</td>\n",
+       "      <td>ciudadano</td>\n",
+       "      <td>ampliar</td>\n",
+       "      <td>y</td>\n",
+       "      <td>confuso</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>0.111046</td>\n",
+       "      <td>uno</td>\n",
+       "      <td>definición</td>\n",
+       "      <td>de</td>\n",
+       "      <td>ciudadano</td>\n",
+       "      <td>ampliar</td>\n",
+       "      <td>y</td>\n",
+       "      <td>confuso</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>4</td>\n",
+       "      <td>0.111574</td>\n",
+       "      <td>uno</td>\n",
+       "      <td>definición</td>\n",
+       "      <td>de</td>\n",
+       "      <td>ciudadano</td>\n",
+       "      <td>ampliar</td>\n",
+       "      <td>y</td>\n",
+       "      <td>confuso</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>5</th>\n",
+       "      <td>5</td>\n",
+       "      <td>0.110600</td>\n",
+       "      <td>uno</td>\n",
+       "      <td>definición</td>\n",
+       "      <td>de</td>\n",
+       "      <td>ciudadano</td>\n",
+       "      <td>ampliar</td>\n",
+       "      <td>y</td>\n",
+       "      <td>confuso</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>6</th>\n",
+       "      <td>6</td>\n",
+       "      <td>0.111129</td>\n",
+       "      <td>uno</td>\n",
+       "      <td>definición</td>\n",
+       "      <td>de</td>\n",
+       "      <td>ciudadano</td>\n",
+       "      <td>ampliar</td>\n",
+       "      <td>y</td>\n",
+       "      <td>confuso</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7</th>\n",
+       "      <td>7</td>\n",
+       "      <td>0.111046</td>\n",
+       "      <td>uno</td>\n",
+       "      <td>definición</td>\n",
+       "      <td>de</td>\n",
+       "      <td>ciudadano</td>\n",
+       "      <td>ampliar</td>\n",
+       "      <td>y</td>\n",
+       "      <td>confuso</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>8</th>\n",
+       "      <td>8</td>\n",
+       "      <td>0.111574</td>\n",
+       "      <td>uno</td>\n",
+       "      <td>definición</td>\n",
+       "      <td>de</td>\n",
+       "      <td>ciudadano</td>\n",
+       "      <td>ampliar</td>\n",
+       "      <td>y</td>\n",
+       "      <td>confuso</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight word1       word2 word3      word4    word5 word6    word7\n",
+       "0     0  0.111574   uno  definición    de  ciudadano  ampliar     y  confuso\n",
+       "1     1  0.110444   uno  definición    de  ciudadano  ampliar     y  confuso\n",
+       "2     2  0.111013   uno  definición    de  ciudadano  ampliar     y  confuso\n",
+       "3     3  0.111046   uno  definición    de  ciudadano  ampliar     y  confuso\n",
+       "4     4  0.111574   uno  definición    de  ciudadano  ampliar     y  confuso\n",
+       "5     5  0.110600   uno  definición    de  ciudadano  ampliar     y  confuso\n",
+       "6     6  0.111129   uno  definición    de  ciudadano  ampliar     y  confuso\n",
+       "7     7  0.111046   uno  definición    de  ciudadano  ampliar     y  confuso\n",
+       "8     8  0.111574   uno  definición    de  ciudadano  ampliar     y  confuso"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.0\n",
+      "model : LinearSVM , score = 0.08862622990678397\n",
+      "model : RBFSVM , score = 0.08862622990678397\n",
+      "model : DecisionTree , score = 0.0\n",
+      "model : RandomForest , score = 0.0\n",
+      "model : AdaBoost , score = 0.0\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 76+75_63+100_40+100_37_63+.csv\n",
+      "Rules(classes) number : 4\n",
+      "Words(features) number : 3\n",
+      "Records number : 5892"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.256020</td>\n",
+       "      <td>no</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>mejorar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.247588</td>\n",
+       "      <td>no</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>mejorar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.248196</td>\n",
+       "      <td>no</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>mejorar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>0.248196</td>\n",
+       "      <td>no</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>mejorar</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight word1  word2    word3\n",
+       "0     0  0.256020    no  haber  mejorar\n",
+       "1     1  0.247588    no  haber  mejorar\n",
+       "2     2  0.248196    no  haber  mejorar\n",
+       "3     3  0.248196    no  haber  mejorar"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.25182717265680205\n",
+      "model : LinearSVM , score = 0.248560018366495\n",
+      "model : RBFSVM , score = 0.2392588726742863\n",
+      "model : DecisionTree , score = 0.22779809088144265\n",
+      "model : RandomForest , score = 0.1955185463236218\n",
+      "model : AdaBoost , score = 0.21712035680478434\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 14+1_29+.csv\n",
+      "Rules(classes) number : 2\n",
+      "Words(features) number : 2\n",
+      "Records number : 12718"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.512582</td>\n",
+       "      <td>forma</td>\n",
+       "      <td>transparente</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.487418</td>\n",
+       "      <td>forma</td>\n",
+       "      <td>transparente</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight  word1         word2\n",
+       "0     0  0.512582  forma  transparente\n",
+       "1     1  0.487418  forma  transparente"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.5107458466229141\n",
+      "model : LinearSVM , score = 0.5118126996132488\n",
+      "model : RBFSVM , score = 0.3961875644243543\n",
+      "model : DecisionTree , score = 0.49368570571413434\n",
+      "model : RandomForest , score = 0.48852175500059875\n",
+      "model : AdaBoost , score = 0.4863975090871166\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 7_29_106_29+8_29_106_29+9_29_106_29+10_108_24+10_108_14_106_29+10_108_1_29_106_29+11_108_24+11_108_14_106_29+11_108_1_29_106_29+33_1_108_24+33_1_108_14_106_29+33_1_108_1_29_106_29+.csv\n",
+      "Rules(classes) number : 12\n",
+      "Words(features) number : 7\n",
+      "Records number : 960"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "      <th>word5</th>\n",
+       "      <th>word6</th>\n",
+       "      <th>word7</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.085027</td>\n",
+       "      <td>el</td>\n",
+       "      <td>comisión</td>\n",
+       "      <td>de</td>\n",
+       "      <td>asunto</td>\n",
+       "      <td>económico</td>\n",
+       "      <td>y</td>\n",
+       "      <td>monetario</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.084131</td>\n",
+       "      <td>el</td>\n",
+       "      <td>comisión</td>\n",
+       "      <td>de</td>\n",
+       "      <td>asunto</td>\n",
+       "      <td>económico</td>\n",
+       "      <td>y</td>\n",
+       "      <td>monetario</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.083727</td>\n",
+       "      <td>el</td>\n",
+       "      <td>comisión</td>\n",
+       "      <td>de</td>\n",
+       "      <td>asunto</td>\n",
+       "      <td>económico</td>\n",
+       "      <td>y</td>\n",
+       "      <td>monetario</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>0.086464</td>\n",
+       "      <td>el</td>\n",
+       "      <td>comisión</td>\n",
+       "      <td>de</td>\n",
+       "      <td>asunto</td>\n",
+       "      <td>económico</td>\n",
+       "      <td>y</td>\n",
+       "      <td>monetario</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>4</td>\n",
+       "      <td>0.079889</td>\n",
+       "      <td>el</td>\n",
+       "      <td>comisión</td>\n",
+       "      <td>de</td>\n",
+       "      <td>asunto</td>\n",
+       "      <td>económico</td>\n",
+       "      <td>y</td>\n",
+       "      <td>monetario</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>5</th>\n",
+       "      <td>5</td>\n",
+       "      <td>0.085027</td>\n",
+       "      <td>el</td>\n",
+       "      <td>comisión</td>\n",
+       "      <td>de</td>\n",
+       "      <td>asunto</td>\n",
+       "      <td>económico</td>\n",
+       "      <td>y</td>\n",
+       "      <td>monetario</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>6</th>\n",
+       "      <td>6</td>\n",
+       "      <td>0.083974</td>\n",
+       "      <td>el</td>\n",
+       "      <td>comisión</td>\n",
+       "      <td>de</td>\n",
+       "      <td>asunto</td>\n",
+       "      <td>económico</td>\n",
+       "      <td>y</td>\n",
+       "      <td>monetario</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7</th>\n",
+       "      <td>7</td>\n",
+       "      <td>0.077760</td>\n",
+       "      <td>el</td>\n",
+       "      <td>comisión</td>\n",
+       "      <td>de</td>\n",
+       "      <td>asunto</td>\n",
+       "      <td>económico</td>\n",
+       "      <td>y</td>\n",
+       "      <td>monetario</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>8</th>\n",
+       "      <td>8</td>\n",
+       "      <td>0.082619</td>\n",
+       "      <td>el</td>\n",
+       "      <td>comisión</td>\n",
+       "      <td>de</td>\n",
+       "      <td>asunto</td>\n",
+       "      <td>económico</td>\n",
+       "      <td>y</td>\n",
+       "      <td>monetario</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>9</th>\n",
+       "      <td>9</td>\n",
+       "      <td>0.086464</td>\n",
+       "      <td>el</td>\n",
+       "      <td>comisión</td>\n",
+       "      <td>de</td>\n",
+       "      <td>asunto</td>\n",
+       "      <td>económico</td>\n",
+       "      <td>y</td>\n",
+       "      <td>monetario</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>10</th>\n",
+       "      <td>10</td>\n",
+       "      <td>0.079889</td>\n",
+       "      <td>el</td>\n",
+       "      <td>comisión</td>\n",
+       "      <td>de</td>\n",
+       "      <td>asunto</td>\n",
+       "      <td>económico</td>\n",
+       "      <td>y</td>\n",
+       "      <td>monetario</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>11</th>\n",
+       "      <td>11</td>\n",
+       "      <td>0.085027</td>\n",
+       "      <td>el</td>\n",
+       "      <td>comisión</td>\n",
+       "      <td>de</td>\n",
+       "      <td>asunto</td>\n",
+       "      <td>económico</td>\n",
+       "      <td>y</td>\n",
+       "      <td>monetario</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "    rule    weight word1     word2 word3   word4      word5 word6      word7\n",
+       "0      0  0.085027    el  comisión    de  asunto  económico     y  monetario\n",
+       "1      1  0.084131    el  comisión    de  asunto  económico     y  monetario\n",
+       "2      2  0.083727    el  comisión    de  asunto  económico     y  monetario\n",
+       "3      3  0.086464    el  comisión    de  asunto  económico     y  monetario\n",
+       "4      4  0.079889    el  comisión    de  asunto  económico     y  monetario\n",
+       "5      5  0.085027    el  comisión    de  asunto  económico     y  monetario\n",
+       "6      6  0.083974    el  comisión    de  asunto  económico     y  monetario\n",
+       "7      7  0.077760    el  comisión    de  asunto  económico     y  monetario\n",
+       "8      8  0.082619    el  comisión    de  asunto  económico     y  monetario\n",
+       "9      9  0.086464    el  comisión    de  asunto  económico     y  monetario\n",
+       "10    10  0.079889    el  comisión    de  asunto  económico     y  monetario\n",
+       "11    11  0.085027    el  comisión    de  asunto  económico     y  monetario"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.07544499220507477\n",
+      "model : LinearSVM , score = 0.052612650939349755\n",
+      "model : RBFSVM , score = 0.08680020851121312\n",
+      "model : DecisionTree , score = 0.017581560870878046\n",
+      "model : RandomForest , score = 0.01424467609890829\n",
+      "model : AdaBoost , score = 0.03310003280258349\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 18+17_29+31_15+31_14+31_1_29+33_29_15+33_29_14+33_29_1_29+.csv\n",
+      "Rules(classes) number : 8\n",
+      "Words(features) number : 4\n",
+      "Records number : 264"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.126349</td>\n",
+       "      <td>el</td>\n",
+       "      <td>nuevo</td>\n",
+       "      <td>iniciativa</td>\n",
+       "      <td>comunitario</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.126684</td>\n",
+       "      <td>el</td>\n",
+       "      <td>nuevo</td>\n",
+       "      <td>iniciativa</td>\n",
+       "      <td>comunitario</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.120451</td>\n",
+       "      <td>el</td>\n",
+       "      <td>nuevo</td>\n",
+       "      <td>iniciativa</td>\n",
+       "      <td>comunitario</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>0.126349</td>\n",
+       "      <td>el</td>\n",
+       "      <td>nuevo</td>\n",
+       "      <td>iniciativa</td>\n",
+       "      <td>comunitario</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>4</td>\n",
+       "      <td>0.126684</td>\n",
+       "      <td>el</td>\n",
+       "      <td>nuevo</td>\n",
+       "      <td>iniciativa</td>\n",
+       "      <td>comunitario</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>5</th>\n",
+       "      <td>5</td>\n",
+       "      <td>0.120451</td>\n",
+       "      <td>el</td>\n",
+       "      <td>nuevo</td>\n",
+       "      <td>iniciativa</td>\n",
+       "      <td>comunitario</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>6</th>\n",
+       "      <td>6</td>\n",
+       "      <td>0.126349</td>\n",
+       "      <td>el</td>\n",
+       "      <td>nuevo</td>\n",
+       "      <td>iniciativa</td>\n",
+       "      <td>comunitario</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7</th>\n",
+       "      <td>7</td>\n",
+       "      <td>0.126684</td>\n",
+       "      <td>el</td>\n",
+       "      <td>nuevo</td>\n",
+       "      <td>iniciativa</td>\n",
+       "      <td>comunitario</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight word1  word2       word3        word4\n",
+       "0     0  0.126349    el  nuevo  iniciativa  comunitario\n",
+       "1     1  0.126684    el  nuevo  iniciativa  comunitario\n",
+       "2     2  0.120451    el  nuevo  iniciativa  comunitario\n",
+       "3     3  0.126349    el  nuevo  iniciativa  comunitario\n",
+       "4     4  0.126684    el  nuevo  iniciativa  comunitario\n",
+       "5     5  0.120451    el  nuevo  iniciativa  comunitario\n",
+       "6     6  0.126349    el  nuevo  iniciativa  comunitario\n",
+       "7     7  0.126684    el  nuevo  iniciativa  comunitario"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.05998969026970071\n",
+      "model : LinearSVM , score = 0.10552628754472712\n",
+      "model : RBFSVM , score = 0.12329034588993983\n",
+      "model : DecisionTree , score = 0.00795886992053042\n",
+      "model : RandomForest , score = 0.0\n",
+      "model : AdaBoost , score = 0.05331442712214913\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 67_63+63_65+63_64_63+63_108_40+63_108_68_63+.csv\n",
+      "Rules(classes) number : 5\n",
+      "Words(features) number : 4\n",
+      "Records number : 10"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.205494</td>\n",
+       "      <td>culpar</td>\n",
+       "      <td>por</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>permitir</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.205494</td>\n",
+       "      <td>culpar</td>\n",
+       "      <td>por</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>permitir</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.205494</td>\n",
+       "      <td>culpar</td>\n",
+       "      <td>por</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>permitir</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>0.191759</td>\n",
+       "      <td>culpar</td>\n",
+       "      <td>por</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>permitir</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>4</td>\n",
+       "      <td>0.191759</td>\n",
+       "      <td>culpar</td>\n",
+       "      <td>por</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>permitir</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight   word1 word2  word3     word4\n",
+       "0     0  0.205494  culpar   por  haber  permitir\n",
+       "1     1  0.205494  culpar   por  haber  permitir\n",
+       "2     2  0.205494  culpar   por  haber  permitir\n",
+       "3     3  0.191759  culpar   por  haber  permitir\n",
+       "4     4  0.191759  culpar   por  haber  permitir"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.205494\n",
+      "model : LinearSVM , score = 0.205494\n",
+      "model : RBFSVM , score = 0.205494\n",
+      "model : DecisionTree , score = 0.205494\n",
+      "model : RandomForest , score = 0.205494\n",
+      "model : AdaBoost , score = 0.205494\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 95+98_108_113+33_96+33_99_108_113+33_113_108_1_108_113+.csv\n",
+      "Rules(classes) number : 5\n",
+      "Words(features) number : 6\n",
+      "Records number : 310"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "/home/aboelhamd/anaconda3/lib/python3.7/site-packages/sklearn/naive_bayes.py:436: RuntimeWarning: divide by zero encountered in log\n",
+      "  n_ij = - 0.5 * np.sum(np.log(2. * np.pi * self.sigma_[i, :]))\n",
+      "/home/aboelhamd/anaconda3/lib/python3.7/site-packages/sklearn/naive_bayes.py:438: RuntimeWarning: invalid value encountered in true_divide\n",
+      "  (self.sigma_[i, :]), 1)\n"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "      <th>word5</th>\n",
+       "      <th>word6</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.213448</td>\n",
+       "      <td>el</td>\n",
+       "      <td>4</td>\n",
+       "      <td>de</td>\n",
+       "      <td>noviembre</td>\n",
+       "      <td>de</td>\n",
+       "      <td>1998</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.201655</td>\n",
+       "      <td>el</td>\n",
+       "      <td>4</td>\n",
+       "      <td>de</td>\n",
+       "      <td>noviembre</td>\n",
+       "      <td>de</td>\n",
+       "      <td>1998</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.205052</td>\n",
+       "      <td>el</td>\n",
+       "      <td>4</td>\n",
+       "      <td>de</td>\n",
+       "      <td>noviembre</td>\n",
+       "      <td>de</td>\n",
+       "      <td>1998</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>0.194317</td>\n",
+       "      <td>el</td>\n",
+       "      <td>4</td>\n",
+       "      <td>de</td>\n",
+       "      <td>noviembre</td>\n",
+       "      <td>de</td>\n",
+       "      <td>1998</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>4</td>\n",
+       "      <td>0.185529</td>\n",
+       "      <td>el</td>\n",
+       "      <td>4</td>\n",
+       "      <td>de</td>\n",
+       "      <td>noviembre</td>\n",
+       "      <td>de</td>\n",
+       "      <td>1998</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight word1  word2 word3      word4 word5  word6\n",
+       "0     0  0.213448    el      4    de  noviembre    de   1998\n",
+       "1     1  0.201655    el      4    de  noviembre    de   1998\n",
+       "2     2  0.205052    el      4    de  noviembre    de   1998\n",
+       "3     3  0.194317    el      4    de  noviembre    de   1998\n",
+       "4     4  0.185529    el      4    de  noviembre    de   1998"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.14465495993573385\n",
+      "model : LinearSVM , score = 0.17372084213541714\n",
+      "model : RBFSVM , score = 0.22295327450479063\n",
+      "model : DecisionTree , score = 0.06054547157295547\n",
+      "model : RandomForest , score = 0.028428275938724212\n",
+      "model : AdaBoost , score = 0.12970190060199288\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 30+100_29+.csv\n",
+      "Rules(classes) number : 2\n",
+      "Words(features) number : 2\n",
+      "Records number : 5240"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.5</td>\n",
+       "      <td>a_escala</td>\n",
+       "      <td>local</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.5</td>\n",
+       "      <td>a_escala</td>\n",
+       "      <td>local</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule  weight     word1  word2\n",
+       "0     0     0.5  a_escala  local\n",
+       "1     1     0.5  a_escala  local"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.48398804800554773\n",
+      "model : LinearSVM , score = 0.4862776422770272\n",
+      "model : RBFSVM , score = 0.3250500585555513\n",
+      "model : DecisionTree , score = 0.465989043359402\n",
+      "model : RandomForest , score = 0.4394429154941928\n",
+      "model : AdaBoost , score = 0.45410508837298397\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 28+16_106_30+16_106_110_29+10_29_106_30+10_29_106_110_29+11_29_106_30+11_29_106_110_29+33_14_106_30+33_14_106_110_29+33_1_29_106_30+33_1_29_106_110_29+.csv\n",
+      "Rules(classes) number : 11\n",
+      "Words(features) number : 6\n",
+      "Records number : 88"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "      <th>word5</th>\n",
+       "      <th>word6</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.096350</td>\n",
+       "      <td>uno</td>\n",
+       "      <td>punto</td>\n",
+       "      <td>concreto</td>\n",
+       "      <td>pero</td>\n",
+       "      <td>muy</td>\n",
+       "      <td>sensible</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.092374</td>\n",
+       "      <td>uno</td>\n",
+       "      <td>punto</td>\n",
+       "      <td>concreto</td>\n",
+       "      <td>pero</td>\n",
+       "      <td>muy</td>\n",
+       "      <td>sensible</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.092374</td>\n",
+       "      <td>uno</td>\n",
+       "      <td>punto</td>\n",
+       "      <td>concreto</td>\n",
+       "      <td>pero</td>\n",
+       "      <td>muy</td>\n",
+       "      <td>sensible</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>0.087653</td>\n",
+       "      <td>uno</td>\n",
+       "      <td>punto</td>\n",
+       "      <td>concreto</td>\n",
+       "      <td>pero</td>\n",
+       "      <td>muy</td>\n",
+       "      <td>sensible</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>4</td>\n",
+       "      <td>0.087653</td>\n",
+       "      <td>uno</td>\n",
+       "      <td>punto</td>\n",
+       "      <td>concreto</td>\n",
+       "      <td>pero</td>\n",
+       "      <td>muy</td>\n",
+       "      <td>sensible</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>5</th>\n",
+       "      <td>5</td>\n",
+       "      <td>0.091771</td>\n",
+       "      <td>uno</td>\n",
+       "      <td>punto</td>\n",
+       "      <td>concreto</td>\n",
+       "      <td>pero</td>\n",
+       "      <td>muy</td>\n",
+       "      <td>sensible</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>6</th>\n",
+       "      <td>6</td>\n",
+       "      <td>0.091771</td>\n",
+       "      <td>uno</td>\n",
+       "      <td>punto</td>\n",
+       "      <td>concreto</td>\n",
+       "      <td>pero</td>\n",
+       "      <td>muy</td>\n",
+       "      <td>sensible</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7</th>\n",
+       "      <td>7</td>\n",
+       "      <td>0.092374</td>\n",
+       "      <td>uno</td>\n",
+       "      <td>punto</td>\n",
+       "      <td>concreto</td>\n",
+       "      <td>pero</td>\n",
+       "      <td>muy</td>\n",
+       "      <td>sensible</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>8</th>\n",
+       "      <td>8</td>\n",
+       "      <td>0.092374</td>\n",
+       "      <td>uno</td>\n",
+       "      <td>punto</td>\n",
+       "      <td>concreto</td>\n",
+       "      <td>pero</td>\n",
+       "      <td>muy</td>\n",
+       "      <td>sensible</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>9</th>\n",
+       "      <td>9</td>\n",
+       "      <td>0.087653</td>\n",
+       "      <td>uno</td>\n",
+       "      <td>punto</td>\n",
+       "      <td>concreto</td>\n",
+       "      <td>pero</td>\n",
+       "      <td>muy</td>\n",
+       "      <td>sensible</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>10</th>\n",
+       "      <td>10</td>\n",
+       "      <td>0.087653</td>\n",
+       "      <td>uno</td>\n",
+       "      <td>punto</td>\n",
+       "      <td>concreto</td>\n",
+       "      <td>pero</td>\n",
+       "      <td>muy</td>\n",
+       "      <td>sensible</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "    rule    weight word1  word2     word3 word4 word5     word6\n",
+       "0      0  0.096350   uno  punto  concreto  pero   muy  sensible\n",
+       "1      1  0.092374   uno  punto  concreto  pero   muy  sensible\n",
+       "2      2  0.092374   uno  punto  concreto  pero   muy  sensible\n",
+       "3      3  0.087653   uno  punto  concreto  pero   muy  sensible\n",
+       "4      4  0.087653   uno  punto  concreto  pero   muy  sensible\n",
+       "5      5  0.091771   uno  punto  concreto  pero   muy  sensible\n",
+       "6      6  0.091771   uno  punto  concreto  pero   muy  sensible\n",
+       "7      7  0.092374   uno  punto  concreto  pero   muy  sensible\n",
+       "8      8  0.092374   uno  punto  concreto  pero   muy  sensible\n",
+       "9      9  0.087653   uno  punto  concreto  pero   muy  sensible\n",
+       "10    10  0.087653   uno  punto  concreto  pero   muy  sensible"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.0\n",
+      "model : LinearSVM , score = 0.09517323380336981\n",
+      "model : RBFSVM , score = 0.09517323380336981\n",
+      "model : DecisionTree , score = 0.0\n",
+      "model : RandomForest , score = 0.0\n",
+      "model : AdaBoost , score = 0.0\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 10_63_106_29+11_63_106_29+33_24+33_1_63_106_29+.csv\n",
+      "Rules(classes) number : 4\n",
+      "Words(features) number : 5\n",
+      "Records number : 48"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "      <th>word5</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.248919</td>\n",
+       "      <td>uno</td>\n",
+       "      <td>modo</td>\n",
+       "      <td>planificar</td>\n",
+       "      <td>y</td>\n",
+       "      <td>sistemático</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.252398</td>\n",
+       "      <td>uno</td>\n",
+       "      <td>modo</td>\n",
+       "      <td>planificar</td>\n",
+       "      <td>y</td>\n",
+       "      <td>sistemático</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.249763</td>\n",
+       "      <td>uno</td>\n",
+       "      <td>modo</td>\n",
+       "      <td>planificar</td>\n",
+       "      <td>y</td>\n",
+       "      <td>sistemático</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>0.248919</td>\n",
+       "      <td>uno</td>\n",
+       "      <td>modo</td>\n",
+       "      <td>planificar</td>\n",
+       "      <td>y</td>\n",
+       "      <td>sistemático</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight word1 word2       word3 word4        word5\n",
+       "0     0  0.248919   uno  modo  planificar     y  sistemático\n",
+       "1     1  0.252398   uno  modo  planificar     y  sistemático\n",
+       "2     2  0.249763   uno  modo  planificar     y  sistemático\n",
+       "3     3  0.248919   uno  modo  planificar     y  sistemático"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.12742105727778924\n",
+      "model : LinearSVM , score = 0.2592056410137905\n",
+      "model : RBFSVM , score = 0.2592056410137905\n",
+      "model : DecisionTree , score = 0.0\n",
+      "model : RandomForest , score = 0.0\n",
+      "model : AdaBoost , score = 0.04150894525417662\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 40_65+40_64_63+40_91_63+40_108_40+40_108_68_63+35_67_63+35_63_65+35_63_64_63+35_63_91_63+35_63_108_40+35_63_108_68_63+37_67_63+37_63_65+37_63_64_63+37_63_91_63+37_63_108_40+37_63_108_68_63+.csv\n",
+      "Rules(classes) number : 17\n",
+      "Words(features) number : 5\n",
+      "Records number : 34"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "      <th>word5</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.058557</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>decir</td>\n",
+       "      <td>de</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>adoptar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.059888</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>decir</td>\n",
+       "      <td>de</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>adoptar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.058393</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>decir</td>\n",
+       "      <td>de</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>adoptar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>0.057981</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>decir</td>\n",
+       "      <td>de</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>adoptar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>4</td>\n",
+       "      <td>0.057981</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>decir</td>\n",
+       "      <td>de</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>adoptar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>5</th>\n",
+       "      <td>5</td>\n",
+       "      <td>0.058962</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>decir</td>\n",
+       "      <td>de</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>adoptar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>6</th>\n",
+       "      <td>6</td>\n",
+       "      <td>0.057254</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>decir</td>\n",
+       "      <td>de</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>adoptar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7</th>\n",
+       "      <td>7</td>\n",
+       "      <td>0.058962</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>decir</td>\n",
+       "      <td>de</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>adoptar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>8</th>\n",
+       "      <td>8</td>\n",
+       "      <td>0.057597</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>decir</td>\n",
+       "      <td>de</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>adoptar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>9</th>\n",
+       "      <td>9</td>\n",
+       "      <td>0.058962</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>decir</td>\n",
+       "      <td>de</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>adoptar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>10</th>\n",
+       "      <td>10</td>\n",
+       "      <td>0.058962</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>decir</td>\n",
+       "      <td>de</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>adoptar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>11</th>\n",
+       "      <td>11</td>\n",
+       "      <td>0.059888</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>decir</td>\n",
+       "      <td>de</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>adoptar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>12</th>\n",
+       "      <td>12</td>\n",
+       "      <td>0.058557</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>decir</td>\n",
+       "      <td>de</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>adoptar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>13</th>\n",
+       "      <td>13</td>\n",
+       "      <td>0.059888</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>decir</td>\n",
+       "      <td>de</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>adoptar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>14</th>\n",
+       "      <td>14</td>\n",
+       "      <td>0.058393</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>decir</td>\n",
+       "      <td>de</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>adoptar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>15</th>\n",
+       "      <td>15</td>\n",
+       "      <td>0.059888</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>decir</td>\n",
+       "      <td>de</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>adoptar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>16</th>\n",
+       "      <td>16</td>\n",
+       "      <td>0.059888</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>decir</td>\n",
+       "      <td>de</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>adoptar</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "    rule    weight  word1  word2 word3  word4    word5\n",
+       "0      0  0.058557  haber  decir    de  haber  adoptar\n",
+       "1      1  0.059888  haber  decir    de  haber  adoptar\n",
+       "2      2  0.058393  haber  decir    de  haber  adoptar\n",
+       "3      3  0.057981  haber  decir    de  haber  adoptar\n",
+       "4      4  0.057981  haber  decir    de  haber  adoptar\n",
+       "5      5  0.058962  haber  decir    de  haber  adoptar\n",
+       "6      6  0.057254  haber  decir    de  haber  adoptar\n",
+       "7      7  0.058962  haber  decir    de  haber  adoptar\n",
+       "8      8  0.057597  haber  decir    de  haber  adoptar\n",
+       "9      9  0.058962  haber  decir    de  haber  adoptar\n",
+       "10    10  0.058962  haber  decir    de  haber  adoptar\n",
+       "11    11  0.059888  haber  decir    de  haber  adoptar\n",
+       "12    12  0.058557  haber  decir    de  haber  adoptar\n",
+       "13    13  0.059888  haber  decir    de  haber  adoptar\n",
+       "14    14  0.058393  haber  decir    de  haber  adoptar\n",
+       "15    15  0.059888  haber  decir    de  haber  adoptar\n",
+       "16    16  0.059888  haber  decir    de  haber  adoptar"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.05855728243281527\n",
+      "model : LinearSVM , score = 0.0598877820336654\n",
+      "model : RBFSVM , score = 0.0598877820336654\n",
+      "model : DecisionTree , score = 0.0598877820336654\n",
+      "model : RandomForest , score = 0.05839268248219526\n",
+      "model : AdaBoost , score = 0.0598877820336654\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 23_29+24_14+24_1_29+14_106_29_14+14_106_29_1_29+1_29_106_29_14+1_29_106_29_1_29+.csv\n",
+      "Rules(classes) number : 7\n",
+      "Words(features) number : 6\n",
+      "Records number : 14"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "/home/aboelhamd/anaconda3/lib/python3.7/site-packages/sklearn/naive_bayes.py:436: RuntimeWarning: divide by zero encountered in log\n",
+      "  n_ij = - 0.5 * np.sum(np.log(2. * np.pi * self.sigma_[i, :]))\n",
+      "/home/aboelhamd/anaconda3/lib/python3.7/site-packages/sklearn/naive_bayes.py:438: RuntimeWarning: invalid value encountered in true_divide\n",
+      "  (self.sigma_[i, :]), 1)\n"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "      <th>word5</th>\n",
+       "      <th>word6</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.143889</td>\n",
+       "      <td>reducción</td>\n",
+       "      <td>obligatorio</td>\n",
+       "      <td>y</td>\n",
+       "      <td>opcional</td>\n",
+       "      <td>propuesta</td>\n",
+       "      <td>relativo</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.144008</td>\n",
+       "      <td>reducción</td>\n",
+       "      <td>obligatorio</td>\n",
+       "      <td>y</td>\n",
+       "      <td>opcional</td>\n",
+       "      <td>propuesta</td>\n",
+       "      <td>relativo</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.143889</td>\n",
+       "      <td>reducción</td>\n",
+       "      <td>obligatorio</td>\n",
+       "      <td>y</td>\n",
+       "      <td>opcional</td>\n",
+       "      <td>propuesta</td>\n",
+       "      <td>relativo</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>0.141088</td>\n",
+       "      <td>reducción</td>\n",
+       "      <td>obligatorio</td>\n",
+       "      <td>y</td>\n",
+       "      <td>opcional</td>\n",
+       "      <td>propuesta</td>\n",
+       "      <td>relativo</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>4</td>\n",
+       "      <td>0.140978</td>\n",
+       "      <td>reducción</td>\n",
+       "      <td>obligatorio</td>\n",
+       "      <td>y</td>\n",
+       "      <td>opcional</td>\n",
+       "      <td>propuesta</td>\n",
+       "      <td>relativo</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>5</th>\n",
+       "      <td>5</td>\n",
+       "      <td>0.143134</td>\n",
+       "      <td>reducción</td>\n",
+       "      <td>obligatorio</td>\n",
+       "      <td>y</td>\n",
+       "      <td>opcional</td>\n",
+       "      <td>propuesta</td>\n",
+       "      <td>relativo</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>6</th>\n",
+       "      <td>6</td>\n",
+       "      <td>0.143015</td>\n",
+       "      <td>reducción</td>\n",
+       "      <td>obligatorio</td>\n",
+       "      <td>y</td>\n",
+       "      <td>opcional</td>\n",
+       "      <td>propuesta</td>\n",
+       "      <td>relativo</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight      word1        word2 word3     word4      word5     word6\n",
+       "0     0  0.143889  reducción  obligatorio     y  opcional  propuesta  relativo\n",
+       "1     1  0.144008  reducción  obligatorio     y  opcional  propuesta  relativo\n",
+       "2     2  0.143889  reducción  obligatorio     y  opcional  propuesta  relativo\n",
+       "3     3  0.141088  reducción  obligatorio     y  opcional  propuesta  relativo\n",
+       "4     4  0.140978  reducción  obligatorio     y  opcional  propuesta  relativo\n",
+       "5     5  0.143134  reducción  obligatorio     y  opcional  propuesta  relativo\n",
+       "6     6  0.143015  reducción  obligatorio     y  opcional  propuesta  relativo"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.1438888561111439\n",
+      "model : LinearSVM , score = 0.14400785599214402\n",
+      "model : RBFSVM , score = 0.14400785599214402\n",
+      "model : DecisionTree , score = 0.14400785599214402\n",
+      "model : RandomForest , score = 0.14400785599214402\n",
+      "model : AdaBoost ,"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "/home/aboelhamd/anaconda3/lib/python3.7/site-packages/sklearn/naive_bayes.py:436: RuntimeWarning: divide by zero encountered in log\n",
+      "  n_ij = - 0.5 * np.sum(np.log(2. * np.pi * self.sigma_[i, :]))\n",
+      "/home/aboelhamd/anaconda3/lib/python3.7/site-packages/sklearn/naive_bayes.py:438: RuntimeWarning: invalid value encountered in true_divide\n",
+      "  (self.sigma_[i, :]), 1)\n"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      " score = 0.14400785599214402\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 10+11+33_1+.csv\n",
+      "Rules(classes) number : 3\n",
+      "Words(features) number : 2\n",
+      "Records number : 269157"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.331459</td>\n",
+       "      <td>el</td>\n",
+       "      <td>escalera</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.337082</td>\n",
+       "      <td>el</td>\n",
+       "      <td>escalera</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.331459</td>\n",
+       "      <td>el</td>\n",
+       "      <td>escalera</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight word1     word2\n",
+       "0     0  0.331459    el  escalera\n",
+       "1     1  0.337082    el  escalera\n",
+       "2     2  0.331459    el  escalera"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.3333842516772396\n",
+      "model : LinearSVM , score = 0.3335066744992964\n",
+      "model : RBFSVM , score = 0.2821422374759117\n",
+      "model : DecisionTree , score = 0.33131072497438363\n",
+      "model : RandomForest , score = 0.3289805767186921\n",
+      "model : AdaBoost , score = 0.3321084830109897\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 76_108_52+75_63_108_52+100_40_108_52+100_37_63_108_52+.csv\n",
+      "Rules(classes) number : 4\n",
+      "Words(features) number : 6\n",
+      "Records number : 8"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "      <th>word5</th>\n",
+       "      <th>word6</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.25</td>\n",
+       "      <td>ya</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>empezar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>hacer</td>\n",
+       "      <td>lo</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.25</td>\n",
+       "      <td>ya</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>empezar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>hacer</td>\n",
+       "      <td>lo</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.25</td>\n",
+       "      <td>ya</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>empezar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>hacer</td>\n",
+       "      <td>lo</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>0.25</td>\n",
+       "      <td>ya</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>empezar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>hacer</td>\n",
+       "      <td>lo</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule  weight word1  word2    word3 word4  word5 word6\n",
+       "0     0    0.25    ya  haber  empezar     a  hacer    lo\n",
+       "1     1    0.25    ya  haber  empezar     a  hacer    lo\n",
+       "2     2    0.25    ya  haber  empezar     a  hacer    lo\n",
+       "3     3    0.25    ya  haber  empezar     a  hacer    lo"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.25\n",
+      "model : LinearSVM , score = 0.25\n",
+      "model : RBFSVM , score = 0.25\n",
+      "model : DecisionTree , score = 0.25\n",
+      "model : RandomForest , score = 0.25\n",
+      "model : AdaBoost , score = 0.25\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 23+24_1+14_106_29_1+1_29_106_29_1+.csv\n",
+      "Rules(classes) number : 4\n",
+      "Words(features) number : 5\n",
+      "Records number : 28"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "/home/aboelhamd/anaconda3/lib/python3.7/site-packages/sklearn/naive_bayes.py:436: RuntimeWarning: divide by zero encountered in log\n",
+      "  n_ij = - 0.5 * np.sum(np.log(2. * np.pi * self.sigma_[i, :]))\n",
+      "/home/aboelhamd/anaconda3/lib/python3.7/site-packages/sklearn/naive_bayes.py:438: RuntimeWarning: invalid value encountered in true_divide\n",
+      "  (self.sigma_[i, :]), 1)\n"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "      <th>word5</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.252352</td>\n",
+       "      <td>consecuencia</td>\n",
+       "      <td>económico</td>\n",
+       "      <td>y</td>\n",
+       "      <td>incalculable</td>\n",
+       "      <td>daño</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.252352</td>\n",
+       "      <td>consecuencia</td>\n",
+       "      <td>económico</td>\n",
+       "      <td>y</td>\n",
+       "      <td>incalculable</td>\n",
+       "      <td>daño</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.251366</td>\n",
+       "      <td>consecuencia</td>\n",
+       "      <td>económico</td>\n",
+       "      <td>y</td>\n",
+       "      <td>incalculable</td>\n",
+       "      <td>daño</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>0.243930</td>\n",
+       "      <td>consecuencia</td>\n",
+       "      <td>económico</td>\n",
+       "      <td>y</td>\n",
+       "      <td>incalculable</td>\n",
+       "      <td>daño</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight         word1      word2 word3         word4 word5\n",
+       "0     0  0.252352  consecuencia  económico     y  incalculable  daño\n",
+       "1     1  0.252352  consecuencia  económico     y  incalculable  daño\n",
+       "2     2  0.251366  consecuencia  económico     y  incalculable  daño\n",
+       "3     3  0.243930  consecuencia  económico     y  incalculable  daño"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.06864049718985643\n",
+      "model : LinearSVM , score = 0.2204557855078027\n",
+      "model : RBFSVM , score = 0.2204557855078027\n",
+      "model : DecisionTree , score = 0.0\n",
+      "model : RandomForest , score = 0.0\n",
+      "model : AdaBoost , score = 0.06763873369214851\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 75_69+100_70+100_37_69+.csv\n",
+      "Rules(classes) number : 3\n",
+      "Words(features) number : 3\n",
+      "Records number : 552"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.316646</td>\n",
+       "      <td>actualmente</td>\n",
+       "      <td>estar</td>\n",
+       "      <td>llevar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.366708</td>\n",
+       "      <td>actualmente</td>\n",
+       "      <td>estar</td>\n",
+       "      <td>llevar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.316646</td>\n",
+       "      <td>actualmente</td>\n",
+       "      <td>estar</td>\n",
+       "      <td>llevar</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight        word1  word2   word3\n",
+       "0     0  0.316646  actualmente  estar  llevar\n",
+       "1     1  0.366708  actualmente  estar  llevar\n",
+       "2     2  0.316646  actualmente  estar  llevar"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.3127688216070106\n",
+      "model : LinearSVM , score = 0.3452572891939757\n",
+      "model : RBFSVM , score = 0.3377194914074114\n",
+      "model : DecisionTree , score = 0.23282676815502515\n",
+      "model : RandomForest , score = 0.2048260357129543\n",
+      "model : AdaBoost , score = 0.21022800798163122\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 40+35_36+35_63+37_36+37_63+.csv\n",
+      "Rules(classes) number : 5\n",
+      "Words(features) number : 2\n",
+      "Records number : 10"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.220455</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>haber_que</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.178101</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>haber_que</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.185300</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>haber_que</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>0.195689</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>haber_que</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>4</td>\n",
+       "      <td>0.220455</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>haber_que</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight  word1      word2\n",
+       "0     0  0.220455  haber  haber_que\n",
+       "1     1  0.178101  haber  haber_que\n",
+       "2     2  0.185300  haber  haber_que\n",
+       "3     3  0.195689  haber  haber_que\n",
+       "4     4  0.220455  haber  haber_que"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.220455\n",
+      "model : LinearSVM , score = 0.220455\n",
+      "model : RBFSVM , score = 0.220455\n",
+      "model : DecisionTree , score = 0.220455\n",
+      "model : RandomForest , score = 0.220455\n",
+      "model : AdaBoost , score = 0.220455\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 7_29+8_29+9_29+10_108_14+10_108_1_29+11_108_14+11_108_1_29+33_1_108_14+33_1_108_1_29+.csv\n",
+      "Rules(classes) number : 9\n",
+      "Words(features) number : 5\n",
+      "Records number : 15057"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "/home/aboelhamd/anaconda3/lib/python3.7/site-packages/sklearn/naive_bayes.py:436: RuntimeWarning: divide by zero encountered in log\n",
+      "  n_ij = - 0.5 * np.sum(np.log(2. * np.pi * self.sigma_[i, :]))\n",
+      "/home/aboelhamd/anaconda3/lib/python3.7/site-packages/sklearn/naive_bayes.py:438: RuntimeWarning: invalid value encountered in true_divide\n",
+      "  (self.sigma_[i, :]), 1)\n"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "      <th>word5</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.110232</td>\n",
+       "      <td>el</td>\n",
+       "      <td>programa</td>\n",
+       "      <td>de</td>\n",
+       "      <td>fondo</td>\n",
+       "      <td>estructural</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.113624</td>\n",
+       "      <td>el</td>\n",
+       "      <td>programa</td>\n",
+       "      <td>de</td>\n",
+       "      <td>fondo</td>\n",
+       "      <td>estructural</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.112062</td>\n",
+       "      <td>el</td>\n",
+       "      <td>programa</td>\n",
+       "      <td>de</td>\n",
+       "      <td>fondo</td>\n",
+       "      <td>estructural</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>0.110865</td>\n",
+       "      <td>el</td>\n",
+       "      <td>programa</td>\n",
+       "      <td>de</td>\n",
+       "      <td>fondo</td>\n",
+       "      <td>estructural</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>4</td>\n",
+       "      <td>0.110232</td>\n",
+       "      <td>el</td>\n",
+       "      <td>programa</td>\n",
+       "      <td>de</td>\n",
+       "      <td>fondo</td>\n",
+       "      <td>estructural</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>5</th>\n",
+       "      <td>5</td>\n",
+       "      <td>0.111261</td>\n",
+       "      <td>el</td>\n",
+       "      <td>programa</td>\n",
+       "      <td>de</td>\n",
+       "      <td>fondo</td>\n",
+       "      <td>estructural</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>6</th>\n",
+       "      <td>6</td>\n",
+       "      <td>0.110628</td>\n",
+       "      <td>el</td>\n",
+       "      <td>programa</td>\n",
+       "      <td>de</td>\n",
+       "      <td>fondo</td>\n",
+       "      <td>estructural</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7</th>\n",
+       "      <td>7</td>\n",
+       "      <td>0.110865</td>\n",
+       "      <td>el</td>\n",
+       "      <td>programa</td>\n",
+       "      <td>de</td>\n",
+       "      <td>fondo</td>\n",
+       "      <td>estructural</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>8</th>\n",
+       "      <td>8</td>\n",
+       "      <td>0.110232</td>\n",
+       "      <td>el</td>\n",
+       "      <td>programa</td>\n",
+       "      <td>de</td>\n",
+       "      <td>fondo</td>\n",
+       "      <td>estructural</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight word1     word2 word3  word4        word5\n",
+       "0     0  0.110232    el  programa    de  fondo  estructural\n",
+       "1     1  0.113624    el  programa    de  fondo  estructural\n",
+       "2     2  0.112062    el  programa    de  fondo  estructural\n",
+       "3     3  0.110865    el  programa    de  fondo  estructural\n",
+       "4     4  0.110232    el  programa    de  fondo  estructural\n",
+       "5     5  0.111261    el  programa    de  fondo  estructural\n",
+       "6     6  0.110628    el  programa    de  fondo  estructural\n",
+       "7     7  0.110865    el  programa    de  fondo  estructural\n",
+       "8     8  0.110232    el  programa    de  fondo  estructural"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.10341161383475868\n",
+      "model : LinearSVM , score = 0.10728662965308675\n",
+      "model : RBFSVM , score = 0.11419949384647897\n",
+      "model : DecisionTree , score = 0.09967279951894004\n",
+      "model : RandomForest , score = 0.08666607275524917\n",
+      "model : AdaBoost , score = 0.0893746415611208\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 7_63_29+8_63_29+9_63_29+10_108_15+10_108_1_63_29+11_108_15+11_108_1_63_29+33_1_108_15+33_1_108_1_63_29+.csv\n",
+      "Rules(classes) number : 9\n",
+      "Words(features) number : 6\n",
+      "Records number : 18"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "      <th>word5</th>\n",
+       "      <th>word6</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.113506</td>\n",
+       "      <td>el</td>\n",
+       "      <td>tribunal</td>\n",
+       "      <td>de</td>\n",
+       "      <td>cuenta</td>\n",
+       "      <td>titular</td>\n",
+       "      <td>político</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.107992</td>\n",
+       "      <td>el</td>\n",
+       "      <td>tribunal</td>\n",
+       "      <td>de</td>\n",
+       "      <td>cuenta</td>\n",
+       "      <td>titular</td>\n",
+       "      <td>político</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.108196</td>\n",
+       "      <td>el</td>\n",
+       "      <td>tribunal</td>\n",
+       "      <td>de</td>\n",
+       "      <td>cuenta</td>\n",
+       "      <td>titular</td>\n",
+       "      <td>político</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>0.110663</td>\n",
+       "      <td>el</td>\n",
+       "      <td>tribunal</td>\n",
+       "      <td>de</td>\n",
+       "      <td>cuenta</td>\n",
+       "      <td>titular</td>\n",
+       "      <td>político</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>4</td>\n",
+       "      <td>0.113506</td>\n",
+       "      <td>el</td>\n",
+       "      <td>tribunal</td>\n",
+       "      <td>de</td>\n",
+       "      <td>cuenta</td>\n",
+       "      <td>titular</td>\n",
+       "      <td>político</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>5</th>\n",
+       "      <td>5</td>\n",
+       "      <td>0.110113</td>\n",
+       "      <td>el</td>\n",
+       "      <td>tribunal</td>\n",
+       "      <td>de</td>\n",
+       "      <td>cuenta</td>\n",
+       "      <td>titular</td>\n",
+       "      <td>político</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>6</th>\n",
+       "      <td>6</td>\n",
+       "      <td>0.111856</td>\n",
+       "      <td>el</td>\n",
+       "      <td>tribunal</td>\n",
+       "      <td>de</td>\n",
+       "      <td>cuenta</td>\n",
+       "      <td>titular</td>\n",
+       "      <td>político</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7</th>\n",
+       "      <td>7</td>\n",
+       "      <td>0.110663</td>\n",
+       "      <td>el</td>\n",
+       "      <td>tribunal</td>\n",
+       "      <td>de</td>\n",
+       "      <td>cuenta</td>\n",
+       "      <td>titular</td>\n",
+       "      <td>político</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>8</th>\n",
+       "      <td>8</td>\n",
+       "      <td>0.113506</td>\n",
+       "      <td>el</td>\n",
+       "      <td>tribunal</td>\n",
+       "      <td>de</td>\n",
+       "      <td>cuenta</td>\n",
+       "      <td>titular</td>\n",
+       "      <td>político</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight word1     word2 word3   word4    word5     word6\n",
+       "0     0  0.113506    el  tribunal    de  cuenta  titular  político\n",
+       "1     1  0.107992    el  tribunal    de  cuenta  titular  político\n",
+       "2     2  0.108196    el  tribunal    de  cuenta  titular  político\n",
+       "3     3  0.110663    el  tribunal    de  cuenta  titular  político\n",
+       "4     4  0.113506    el  tribunal    de  cuenta  titular  político\n",
+       "5     5  0.110113    el  tribunal    de  cuenta  titular  político\n",
+       "6     6  0.111856    el  tribunal    de  cuenta  titular  político\n",
+       "7     7  0.110663    el  tribunal    de  cuenta  titular  político\n",
+       "8     8  0.113506    el  tribunal    de  cuenta  titular  político"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.11350588649411349\n",
+      "model : LinearSVM , score = 0.11350588649411349\n",
+      "model : RBFSVM , score = 0.11350588649411349\n",
+      "model : DecisionTree , score = 0.11350588649411349\n",
+      "model : RandomForest , score = 0.11066288933711066\n",
+      "model : AdaBoost , score = 0.11350588649411349\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "/home/aboelhamd/anaconda3/lib/python3.7/site-packages/sklearn/naive_bayes.py:436: RuntimeWarning: divide by zero encountered in log\n",
+      "  n_ij = - 0.5 * np.sum(np.log(2. * np.pi * self.sigma_[i, :]))\n",
+      "/home/aboelhamd/anaconda3/lib/python3.7/site-packages/sklearn/naive_bayes.py:438: RuntimeWarning: invalid value encountered in true_divide\n",
+      "  (self.sigma_[i, :]), 1)\n"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 40_108_52+37_63_108_52+.csv\n",
+      "Rules(classes) number : 2\n",
+      "Words(features) number : 5\n",
+      "Records number : 6"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "      <th>word5</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.5</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>esforzar</td>\n",
+       "      <td>por</td>\n",
+       "      <td>lograr</td>\n",
+       "      <td>lo</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.5</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>esforzar</td>\n",
+       "      <td>por</td>\n",
+       "      <td>lograr</td>\n",
+       "      <td>lo</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule  weight  word1     word2 word3   word4 word5\n",
+       "0     0     0.5  haber  esforzar   por  lograr    lo\n",
+       "1     1     0.5  haber  esforzar   por  lograr    lo"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.3333333333333333\n",
+      "model : LinearSVM , score = 0.3333333333333333\n",
+      "model : RBFSVM , score = 0.3333333333333333\n",
+      "model : DecisionTree , score = 0.3333333333333333\n",
+      "model : RandomForest , score = 0.3333333333333333\n",
+      "model : AdaBoost , score = 0.3333333333333333\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 75_108_52+100_37_108_52+.csv\n",
+      "Rules(classes) number : 2\n",
+      "Words(features) number : 5\n",
+      "Records number : 4"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "      <th>word5</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.524038</td>\n",
+       "      <td>no</td>\n",
+       "      <td>llegar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>hacer</td>\n",
+       "      <td>lo</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.475962</td>\n",
+       "      <td>no</td>\n",
+       "      <td>llegar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>hacer</td>\n",
+       "      <td>lo</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight word1   word2 word3  word4 word5\n",
+       "0     0  0.524038    no  llegar     a  hacer    lo\n",
+       "1     1  0.475962    no  llegar     a  hacer    lo"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.524038\n",
+      "model : LinearSVM , score = 0.524038\n",
+      "model : RBFSVM , score = 0.524038\n",
+      "model : DecisionTree , score = 0.524038\n",
+      "model : RandomForest , score = 0.524038\n",
+      "model : AdaBoost , score = 0.524038\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 22+12_30+12_110_29+113_1_30+113_1_110_29+.csv\n",
+      "Rules(classes) number : 5\n",
+      "Words(features) number : 4\n",
+      "Records number : 55"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "/home/aboelhamd/anaconda3/lib/python3.7/site-packages/sklearn/naive_bayes.py:436: RuntimeWarning: divide by zero encountered in log\n",
+      "  n_ij = - 0.5 * np.sum(np.log(2. * np.pi * self.sigma_[i, :]))\n",
+      "/home/aboelhamd/anaconda3/lib/python3.7/site-packages/sklearn/naive_bayes.py:438: RuntimeWarning: invalid value encountered in true_divide\n",
+      "  (self.sigma_[i, :]), 1)\n"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.196140</td>\n",
+       "      <td>tres</td>\n",
+       "      <td>pregunta</td>\n",
+       "      <td>muy</td>\n",
+       "      <td>amplio</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.200965</td>\n",
+       "      <td>tres</td>\n",
+       "      <td>pregunta</td>\n",
+       "      <td>muy</td>\n",
+       "      <td>amplio</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.200965</td>\n",
+       "      <td>tres</td>\n",
+       "      <td>pregunta</td>\n",
+       "      <td>muy</td>\n",
+       "      <td>amplio</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>0.200965</td>\n",
+       "      <td>tres</td>\n",
+       "      <td>pregunta</td>\n",
+       "      <td>muy</td>\n",
+       "      <td>amplio</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>4</td>\n",
+       "      <td>0.200965</td>\n",
+       "      <td>tres</td>\n",
+       "      <td>pregunta</td>\n",
+       "      <td>muy</td>\n",
+       "      <td>amplio</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight word1     word2 word3   word4\n",
+       "0     0  0.196140  tres  pregunta   muy  amplio\n",
+       "1     1  0.200965  tres  pregunta   muy  amplio\n",
+       "2     2  0.200965  tres  pregunta   muy  amplio\n",
+       "3     3  0.200965  tres  pregunta   muy  amplio\n",
+       "4     4  0.200965  tres  pregunta   muy  amplio"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.0\n",
+      "model : LinearSVM , score = 0.18761920526265205\n",
+      "model : RBFSVM , score = 0.18761920526265205\n",
+      "model : DecisionTree , score = 0.0\n",
+      "model : RandomForest , score = 0.0\n",
+      "model : AdaBoost , score = 0.10754406011611722\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 21+10_30+10_100_29+11_30+11_100_29+33_1_30+33_1_100_29+.csv\n",
+      "Rules(classes) number : 7\n",
+      "Words(features) number : 4\n",
+      "Records number : 6314"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.139609</td>\n",
+       "      <td>ninguno</td>\n",
+       "      <td>responsable</td>\n",
+       "      <td>claramente</td>\n",
+       "      <td>definido</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.139571</td>\n",
+       "      <td>ninguno</td>\n",
+       "      <td>responsable</td>\n",
+       "      <td>claramente</td>\n",
+       "      <td>definido</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.139571</td>\n",
+       "      <td>ninguno</td>\n",
+       "      <td>responsable</td>\n",
+       "      <td>claramente</td>\n",
+       "      <td>definido</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>0.151054</td>\n",
+       "      <td>ninguno</td>\n",
+       "      <td>responsable</td>\n",
+       "      <td>claramente</td>\n",
+       "      <td>definido</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>4</td>\n",
+       "      <td>0.151054</td>\n",
+       "      <td>ninguno</td>\n",
+       "      <td>responsable</td>\n",
+       "      <td>claramente</td>\n",
+       "      <td>definido</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>5</th>\n",
+       "      <td>5</td>\n",
+       "      <td>0.139571</td>\n",
+       "      <td>ninguno</td>\n",
+       "      <td>responsable</td>\n",
+       "      <td>claramente</td>\n",
+       "      <td>definido</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>6</th>\n",
+       "      <td>6</td>\n",
+       "      <td>0.139571</td>\n",
+       "      <td>ninguno</td>\n",
+       "      <td>responsable</td>\n",
+       "      <td>claramente</td>\n",
+       "      <td>definido</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight    word1        word2       word3     word4\n",
+       "0     0  0.139609  ninguno  responsable  claramente  definido\n",
+       "1     1  0.139571  ninguno  responsable  claramente  definido\n",
+       "2     2  0.139571  ninguno  responsable  claramente  definido\n",
+       "3     3  0.151054  ninguno  responsable  claramente  definido\n",
+       "4     4  0.151054  ninguno  responsable  claramente  definido\n",
+       "5     5  0.139571  ninguno  responsable  claramente  definido\n",
+       "6     6  0.139571  ninguno  responsable  claramente  definido"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.13484570998686124\n",
+      "model : LinearSVM , score = 0.1491420531697449\n",
+      "model : RBFSVM , score = 0.1491420531697449\n",
+      "model : DecisionTree , score = 0.1357829061167114\n",
+      "model : RandomForest , score = 0.1049334948840881\n",
+      "model : AdaBoost , score = 0.11366428395684831\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 30_64+30_91+30_108_68+100_67+100_63_64+100_63_91+100_63_108_68+.csv\n",
+      "Rules(classes) number : 7\n",
+      "Words(features) number : 4\n",
+      "Records number : 14"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.144291</td>\n",
+       "      <td>sólo</td>\n",
+       "      <td>hacer</td>\n",
+       "      <td>de</td>\n",
+       "      <td>ostentar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.142893</td>\n",
+       "      <td>sólo</td>\n",
+       "      <td>hacer</td>\n",
+       "      <td>de</td>\n",
+       "      <td>ostentar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.140671</td>\n",
+       "      <td>sólo</td>\n",
+       "      <td>hacer</td>\n",
+       "      <td>de</td>\n",
+       "      <td>ostentar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>0.144291</td>\n",
+       "      <td>sólo</td>\n",
+       "      <td>hacer</td>\n",
+       "      <td>de</td>\n",
+       "      <td>ostentar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>4</td>\n",
+       "      <td>0.144291</td>\n",
+       "      <td>sólo</td>\n",
+       "      <td>hacer</td>\n",
+       "      <td>de</td>\n",
+       "      <td>ostentar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>5</th>\n",
+       "      <td>5</td>\n",
+       "      <td>0.142893</td>\n",
+       "      <td>sólo</td>\n",
+       "      <td>hacer</td>\n",
+       "      <td>de</td>\n",
+       "      <td>ostentar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>6</th>\n",
+       "      <td>6</td>\n",
+       "      <td>0.140671</td>\n",
+       "      <td>sólo</td>\n",
+       "      <td>hacer</td>\n",
+       "      <td>de</td>\n",
+       "      <td>ostentar</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight word1  word2 word3     word4\n",
+       "0     0  0.144291  sólo  hacer    de  ostentar\n",
+       "1     1  0.142893  sólo  hacer    de  ostentar\n",
+       "2     2  0.140671  sólo  hacer    de  ostentar\n",
+       "3     3  0.144291  sólo  hacer    de  ostentar\n",
+       "4     4  0.144291  sólo  hacer    de  ostentar\n",
+       "5     5  0.142893  sólo  hacer    de  ostentar\n",
+       "6     6  0.140671  sólo  hacer    de  ostentar"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.1442908557091443\n",
+      "model : LinearSVM , score = 0.1442908557091443\n",
+      "model : RBFSVM , score = 0.1442908557091443\n",
+      "model : DecisionTree , score = 0.1442908557091443\n",
+      "model : RandomForest , score = 0.14067085932914067\n",
+      "model : AdaBoost , score = 0.1442908557091443\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 107+106_114+.csv\n",
+      "Rules(classes) number : 2\n",
+      "Words(features) number : 2\n",
+      "Records number : 22"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "/home/aboelhamd/anaconda3/lib/python3.7/site-packages/sklearn/naive_bayes.py:436: RuntimeWarning: divide by zero encountered in log\n",
+      "  n_ij = - 0.5 * np.sum(np.log(2. * np.pi * self.sigma_[i, :]))\n",
+      "/home/aboelhamd/anaconda3/lib/python3.7/site-packages/sklearn/naive_bayes.py:438: RuntimeWarning: invalid value encountered in true_divide\n",
+      "  (self.sigma_[i, :]), 1)\n"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.492222</td>\n",
+       "      <td>por_ello</td>\n",
+       "      <td>?</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.507778</td>\n",
+       "      <td>por_ello</td>\n",
+       "      <td>?</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight     word1 word2\n",
+       "0     0  0.492222  por_ello     ?\n",
+       "1     1  0.507778  por_ello     ?"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.5162341376179551\n",
+      "model : LinearSVM , score = 0.4837658623820448\n",
+      "model : RBFSVM , score = 0.4837658623820448\n",
+      "model : DecisionTree , score = 0.4837658623820448\n",
+      "model : RandomForest , score = 0.4837658623820448\n",
+      "model : AdaBoost ,"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "/home/aboelhamd/anaconda3/lib/python3.7/site-packages/sklearn/naive_bayes.py:436: RuntimeWarning: divide by zero encountered in log\n",
+      "  n_ij = - 0.5 * np.sum(np.log(2. * np.pi * self.sigma_[i, :]))\n",
+      "/home/aboelhamd/anaconda3/lib/python3.7/site-packages/sklearn/naive_bayes.py:438: RuntimeWarning: divide by zero encountered in true_divide\n",
+      "  (self.sigma_[i, :]), 1)\n",
+      "/home/aboelhamd/anaconda3/lib/python3.7/site-packages/sklearn/naive_bayes.py:438: RuntimeWarning: invalid value encountered in true_divide\n",
+      "  (self.sigma_[i, :]), 1)\n"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      " score = 0.4837658623820448\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 66_63+108_33_40+108_33_68_63+.csv\n",
+      "Rules(classes) number : 3\n",
+      "Words(features) number : 4\n",
+      "Records number : 12"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.332287</td>\n",
+       "      <td>a</td>\n",
+       "      <td>el</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>ser</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.333856</td>\n",
+       "      <td>a</td>\n",
+       "      <td>el</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>ser</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.333856</td>\n",
+       "      <td>a</td>\n",
+       "      <td>el</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>ser</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight word1 word2  word3 word4\n",
+       "0     0  0.332287     a    el  haber   ser\n",
+       "1     1  0.333856     a    el  haber   ser\n",
+       "2     2  0.333856     a    el  haber   ser"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.168756715514194\n",
+      "model : LinearSVM , score = 0.3317720257615749\n",
+      "model : RBFSVM , score = 0.3317720257615749\n",
+      "model : DecisionTree , score = 0.168756715514194\n",
+      "model : RandomForest , score = 0.168756715514194\n",
+      "model : AdaBoost , score = 0.168756715514194\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 73+47_69+.csv\n",
+      "Rules(classes) number : 2\n",
+      "Words(features) number : 3\n",
+      "Records number : 20"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.516462</td>\n",
+       "      <td>lo</td>\n",
+       "      <td>estar</td>\n",
+       "      <td>considerar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.483538</td>\n",
+       "      <td>lo</td>\n",
+       "      <td>estar</td>\n",
+       "      <td>considerar</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight word1  word2       word3\n",
+       "0     0  0.516462    lo  estar  considerar\n",
+       "1     1  0.483538    lo  estar  considerar"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.2885738316797356\n",
+      "model : LinearSVM , score = 0.5246475627579833\n",
+      "model : RBFSVM , score = 0.5246475627579833\n",
+      "model : DecisionTree , score = 0.2885738316797356\n",
+      "model : RandomForest , score = 0.2885738316797356\n",
+      "model : AdaBoost , score = 0.2885738316797356\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 76+92_63+100_40+100_68_63+.csv\n",
+      "Rules(classes) number : 4\n",
+      "Words(features) number : 3\n",
+      "Records number : 52"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.254160</td>\n",
+       "      <td>no</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>dar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.251783</td>\n",
+       "      <td>no</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>dar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.247029</td>\n",
+       "      <td>no</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>dar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>0.247029</td>\n",
+       "      <td>no</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>dar</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight word1  word2 word3\n",
+       "0     0  0.254160    no  haber   dar\n",
+       "1     1  0.251783    no  haber   dar\n",
+       "2     2  0.247029    no  haber   dar\n",
+       "3     3  0.247029    no  haber   dar"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.15103283830142405\n",
+      "model : LinearSVM , score = 0.22423955793981423\n",
+      "model : RBFSVM , score = 0.22423955793981423\n",
+      "model : DecisionTree , score = 0.03948810955170136\n",
+      "model : RandomForest , score = 0.037898620798693966\n",
+      "model : AdaBoost , score = 0.03948810955170136\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 67+63_64+63_108_68+.csv\n",
+      "Rules(classes) number : 3\n",
+      "Words(features) number : 3\n",
+      "Records number : 1440"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.333333</td>\n",
+       "      <td>encaminar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>colmar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.333333</td>\n",
+       "      <td>encaminar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>colmar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.333333</td>\n",
+       "      <td>encaminar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>colmar</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight      word1 word2   word3\n",
+       "0     0  0.333333  encaminar     a  colmar\n",
+       "1     1  0.333333  encaminar     a  colmar\n",
+       "2     2  0.333333  encaminar     a  colmar"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.3036674273824966\n",
+      "model : LinearSVM , score = 0.30277620674779554\n",
+      "model : RBFSVM , score = 0.2817275907721134\n",
+      "model : DecisionTree , score = 0.2409576074269052\n",
+      "model : RandomForest , score = 0.2251987337310944\n",
+      "model : AdaBoost , score = 0.2475688472396243\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 108_85+108_100_52+.csv\n",
+      "Rules(classes) number : 2\n",
+      "Words(features) number : 4\n",
+      "Records number : 6"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.504829</td>\n",
+       "      <td>para</td>\n",
+       "      <td>no</td>\n",
+       "      <td>hacer</td>\n",
+       "      <td>lo</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.495171</td>\n",
+       "      <td>para</td>\n",
+       "      <td>no</td>\n",
+       "      <td>hacer</td>\n",
+       "      <td>lo</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight word1 word2  word3 word4\n",
+       "0     0  0.504829  para    no  hacer    lo\n",
+       "1     1  0.495171  para    no  hacer    lo"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.6778889827349154\n",
+      "model : LinearSVM , score = 0.3221110172650846\n",
+      "model : RBFSVM , score = 0.3221110172650846\n",
+      "model : DecisionTree , score = 0.3221110172650846\n",
+      "model : RandomForest , score = 0.3221110172650846\n",
+      "model : AdaBoost , score = 0.3221110172650846\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 75_12+75_113_1+100_13+100_37_12+100_37_113_1+.csv\n",
+      "Rules(classes) number : 5\n",
+      "Words(features) number : 4\n",
+      "Records number : 20"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "/home/aboelhamd/anaconda3/lib/python3.7/site-packages/sklearn/naive_bayes.py:436: RuntimeWarning: divide by zero encountered in log\n",
+      "  n_ij = - 0.5 * np.sum(np.log(2. * np.pi * self.sigma_[i, :]))\n",
+      "/home/aboelhamd/anaconda3/lib/python3.7/site-packages/sklearn/naive_bayes.py:438: RuntimeWarning: invalid value encountered in true_divide\n",
+      "  (self.sigma_[i, :]), 1)\n"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.191526</td>\n",
+       "      <td>allí</td>\n",
+       "      <td>hacer</td>\n",
+       "      <td>dos</td>\n",
+       "      <td>año</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.191526</td>\n",
+       "      <td>allí</td>\n",
+       "      <td>hacer</td>\n",
+       "      <td>dos</td>\n",
+       "      <td>año</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.233895</td>\n",
+       "      <td>allí</td>\n",
+       "      <td>hacer</td>\n",
+       "      <td>dos</td>\n",
+       "      <td>año</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>0.191526</td>\n",
+       "      <td>allí</td>\n",
+       "      <td>hacer</td>\n",
+       "      <td>dos</td>\n",
+       "      <td>año</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>4</td>\n",
+       "      <td>0.191526</td>\n",
+       "      <td>allí</td>\n",
+       "      <td>hacer</td>\n",
+       "      <td>dos</td>\n",
+       "      <td>año</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight word1  word2 word3 word4\n",
+       "0     0  0.191526  allí  hacer   dos   año\n",
+       "1     1  0.191526  allí  hacer   dos   año\n",
+       "2     2  0.233895  allí  hacer   dos   año\n",
+       "3     3  0.191526  allí  hacer   dos   año\n",
+       "4     4  0.191526  allí  hacer   dos   año"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.0\n",
+      "model : LinearSVM , score = 0.20502695568933266\n",
+      "model : RBFSVM , score = 0.20502695568933266\n",
+      "model : DecisionTree , score = 0.0\n",
+      "model : RandomForest , score = 0.0\n",
+      "model : AdaBoost , score = 0.0\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 38+68_64+68_108_68+.csv\n",
+      "Rules(classes) number : 3\n",
+      "Words(features) number : 3\n",
+      "Records number : 48"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.342926</td>\n",
+       "      <td>llegar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>llevar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.328537</td>\n",
+       "      <td>llegar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>llevar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.328537</td>\n",
+       "      <td>llegar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>llevar</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight   word1 word2   word3\n",
+       "0     0  0.342926  llegar     a  llevar\n",
+       "1     1  0.328537  llegar     a  llevar\n",
+       "2     2  0.328537  llegar     a  llevar"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.2800503298865964\n",
+      "model : LinearSVM , score = 0.35847536180810363\n",
+      "model : RBFSVM , score = 0.35847536180810363\n",
+      "model : DecisionTree , score = 0.11863372226489452\n",
+      "model : RandomForest , score = 0.11863372226489452\n",
+      "model : AdaBoost , score = 0.15941830265683726\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 94_1+97_108_12+97_108_113_1+108_95_1+108_98_108_12+108_98_108_113_1+108_33_96_1+108_33_99_108_12+108_33_99_108_113_1+108_33_113_108_1_108_12+108_33_113_108_1_108_113_1+.csv\n",
+      "Rules(classes) number : 11\n",
+      "Words(features) number : 8\n",
+      "Records number : 22"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "      <th>word5</th>\n",
+       "      <th>word6</th>\n",
+       "      <th>word7</th>\n",
+       "      <th>word8</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.099314</td>\n",
+       "      <td>hasta</td>\n",
+       "      <td>el</td>\n",
+       "      <td>3</td>\n",
+       "      <td>de</td>\n",
+       "      <td>julio</td>\n",
+       "      <td>de</td>\n",
+       "      <td>2008</td>\n",
+       "      <td>artículo</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.093227</td>\n",
+       "      <td>hasta</td>\n",
+       "      <td>el</td>\n",
+       "      <td>3</td>\n",
+       "      <td>de</td>\n",
+       "      <td>julio</td>\n",
+       "      <td>de</td>\n",
+       "      <td>2008</td>\n",
+       "      <td>artículo</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.093227</td>\n",
+       "      <td>hasta</td>\n",
+       "      <td>el</td>\n",
+       "      <td>3</td>\n",
+       "      <td>de</td>\n",
+       "      <td>julio</td>\n",
+       "      <td>de</td>\n",
+       "      <td>2008</td>\n",
+       "      <td>artículo</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>0.093704</td>\n",
+       "      <td>hasta</td>\n",
+       "      <td>el</td>\n",
+       "      <td>3</td>\n",
+       "      <td>de</td>\n",
+       "      <td>julio</td>\n",
+       "      <td>de</td>\n",
+       "      <td>2008</td>\n",
+       "      <td>artículo</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>4</td>\n",
+       "      <td>0.088948</td>\n",
+       "      <td>hasta</td>\n",
+       "      <td>el</td>\n",
+       "      <td>3</td>\n",
+       "      <td>de</td>\n",
+       "      <td>julio</td>\n",
+       "      <td>de</td>\n",
+       "      <td>2008</td>\n",
+       "      <td>artículo</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>5</th>\n",
+       "      <td>5</td>\n",
+       "      <td>0.088948</td>\n",
+       "      <td>hasta</td>\n",
+       "      <td>el</td>\n",
+       "      <td>3</td>\n",
+       "      <td>de</td>\n",
+       "      <td>julio</td>\n",
+       "      <td>de</td>\n",
+       "      <td>2008</td>\n",
+       "      <td>artículo</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>6</th>\n",
+       "      <td>6</td>\n",
+       "      <td>0.093521</td>\n",
+       "      <td>hasta</td>\n",
+       "      <td>el</td>\n",
+       "      <td>3</td>\n",
+       "      <td>de</td>\n",
+       "      <td>julio</td>\n",
+       "      <td>de</td>\n",
+       "      <td>2008</td>\n",
+       "      <td>artículo</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7</th>\n",
+       "      <td>7</td>\n",
+       "      <td>0.088774</td>\n",
+       "      <td>hasta</td>\n",
+       "      <td>el</td>\n",
+       "      <td>3</td>\n",
+       "      <td>de</td>\n",
+       "      <td>julio</td>\n",
+       "      <td>de</td>\n",
+       "      <td>2008</td>\n",
+       "      <td>artículo</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>8</th>\n",
+       "      <td>8</td>\n",
+       "      <td>0.088774</td>\n",
+       "      <td>hasta</td>\n",
+       "      <td>el</td>\n",
+       "      <td>3</td>\n",
+       "      <td>de</td>\n",
+       "      <td>julio</td>\n",
+       "      <td>de</td>\n",
+       "      <td>2008</td>\n",
+       "      <td>artículo</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>9</th>\n",
+       "      <td>9</td>\n",
+       "      <td>0.085781</td>\n",
+       "      <td>hasta</td>\n",
+       "      <td>el</td>\n",
+       "      <td>3</td>\n",
+       "      <td>de</td>\n",
+       "      <td>julio</td>\n",
+       "      <td>de</td>\n",
+       "      <td>2008</td>\n",
+       "      <td>artículo</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>10</th>\n",
+       "      <td>10</td>\n",
+       "      <td>0.085781</td>\n",
+       "      <td>hasta</td>\n",
+       "      <td>el</td>\n",
+       "      <td>3</td>\n",
+       "      <td>de</td>\n",
+       "      <td>julio</td>\n",
+       "      <td>de</td>\n",
+       "      <td>2008</td>\n",
+       "      <td>artículo</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "    rule    weight  word1 word2  word3 word4  word5 word6  word7     word8\n",
+       "0      0  0.099314  hasta    el      3    de  julio    de   2008  artículo\n",
+       "1      1  0.093227  hasta    el      3    de  julio    de   2008  artículo\n",
+       "2      2  0.093227  hasta    el      3    de  julio    de   2008  artículo\n",
+       "3      3  0.093704  hasta    el      3    de  julio    de   2008  artículo\n",
+       "4      4  0.088948  hasta    el      3    de  julio    de   2008  artículo\n",
+       "5      5  0.088948  hasta    el      3    de  julio    de   2008  artículo\n",
+       "6      6  0.093521  hasta    el      3    de  julio    de   2008  artículo\n",
+       "7      7  0.088774  hasta    el      3    de  julio    de   2008  artículo\n",
+       "8      8  0.088774  hasta    el      3    de  julio    de   2008  artículo\n",
+       "9      9  0.085781  hasta    el      3    de  julio    de   2008  artículo\n",
+       "10    10  0.085781  hasta    el      3    de  julio    de   2008  artículo"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.0993142\n",
+      "model : LinearSVM , score = 0.0993142\n",
+      "model : RBFSVM , score = 0.0993142\n",
+      "model : DecisionTree , score = 0.0993142\n",
+      "model : RandomForest , score = 0.08894850000000001\n",
+      "model : AdaBoost , score = 0.0993142\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 40_64+40_108_68+37_67+37_63_64+37_63_108_68+.csv\n",
+      "Rules(classes) number : 5\n",
+      "Words(features) number : 4\n",
+      "Records number : 445"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "/home/aboelhamd/anaconda3/lib/python3.7/site-packages/sklearn/naive_bayes.py:436: RuntimeWarning: divide by zero encountered in log\n",
+      "  n_ij = - 0.5 * np.sum(np.log(2. * np.pi * self.sigma_[i, :]))\n",
+      "/home/aboelhamd/anaconda3/lib/python3.7/site-packages/sklearn/naive_bayes.py:438: RuntimeWarning: invalid value encountered in true_divide\n",
+      "  (self.sigma_[i, :]), 1)\n"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.2</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>entrar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>formar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.2</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>entrar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>formar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.2</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>entrar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>formar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>0.2</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>entrar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>formar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>4</td>\n",
+       "      <td>0.2</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>entrar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>formar</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule  weight  word1   word2 word3   word4\n",
+       "0     0     0.2  haber  entrar     a  formar\n",
+       "1     1     0.2  haber  entrar     a  formar\n",
+       "2     2     0.2  haber  entrar     a  formar\n",
+       "3     3     0.2  haber  entrar     a  formar\n",
+       "4     4     0.2  haber  entrar     a  formar"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.15753374623659042\n",
+      "model : LinearSVM , score = 0.17511385970256155\n",
+      "model : RBFSVM , score = 0.19861279595487677\n",
+      "model : DecisionTree , score = 0.053446455252070986\n",
+      "model : RandomForest , score = 0.05822788654192368\n",
+      "model : AdaBoost , score = 0.07641758281833691\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 7_29_29+8_29_29+9_29_29+10_108_15+10_108_14_29+10_108_1_29_29+11_108_15+11_108_14_29+11_108_1_29_29+33_1_108_15+33_1_108_14_29+33_1_108_1_29_29+.csv\n",
+      "Rules(classes) number : 12\n",
+      "Words(features) number : 6\n",
+      "Records number : 708"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "      <th>word5</th>\n",
+       "      <th>word6</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.081911</td>\n",
+       "      <td>el</td>\n",
+       "      <td>convenio</td>\n",
+       "      <td>de</td>\n",
+       "      <td>ginebra</td>\n",
+       "      <td>derecho</td>\n",
+       "      <td>humanitario</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.087933</td>\n",
+       "      <td>el</td>\n",
+       "      <td>convenio</td>\n",
+       "      <td>de</td>\n",
+       "      <td>ginebra</td>\n",
+       "      <td>derecho</td>\n",
+       "      <td>humanitario</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.085374</td>\n",
+       "      <td>el</td>\n",
+       "      <td>convenio</td>\n",
+       "      <td>de</td>\n",
+       "      <td>ginebra</td>\n",
+       "      <td>derecho</td>\n",
+       "      <td>humanitario</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>0.083864</td>\n",
+       "      <td>el</td>\n",
+       "      <td>convenio</td>\n",
+       "      <td>de</td>\n",
+       "      <td>ginebra</td>\n",
+       "      <td>derecho</td>\n",
+       "      <td>humanitario</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>4</td>\n",
+       "      <td>0.083349</td>\n",
+       "      <td>el</td>\n",
+       "      <td>convenio</td>\n",
+       "      <td>de</td>\n",
+       "      <td>ginebra</td>\n",
+       "      <td>derecho</td>\n",
+       "      <td>humanitario</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>5</th>\n",
+       "      <td>5</td>\n",
+       "      <td>0.081911</td>\n",
+       "      <td>el</td>\n",
+       "      <td>convenio</td>\n",
+       "      <td>de</td>\n",
+       "      <td>ginebra</td>\n",
+       "      <td>derecho</td>\n",
+       "      <td>humanitario</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>6</th>\n",
+       "      <td>6</td>\n",
+       "      <td>0.082985</td>\n",
+       "      <td>el</td>\n",
+       "      <td>convenio</td>\n",
+       "      <td>de</td>\n",
+       "      <td>ginebra</td>\n",
+       "      <td>derecho</td>\n",
+       "      <td>humanitario</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7</th>\n",
+       "      <td>7</td>\n",
+       "      <td>0.082480</td>\n",
+       "      <td>el</td>\n",
+       "      <td>convenio</td>\n",
+       "      <td>de</td>\n",
+       "      <td>ginebra</td>\n",
+       "      <td>derecho</td>\n",
+       "      <td>humanitario</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>8</th>\n",
+       "      <td>8</td>\n",
+       "      <td>0.081071</td>\n",
+       "      <td>el</td>\n",
+       "      <td>convenio</td>\n",
+       "      <td>de</td>\n",
+       "      <td>ginebra</td>\n",
+       "      <td>derecho</td>\n",
+       "      <td>humanitario</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>9</th>\n",
+       "      <td>9</td>\n",
+       "      <td>0.083864</td>\n",
+       "      <td>el</td>\n",
+       "      <td>convenio</td>\n",
+       "      <td>de</td>\n",
+       "      <td>ginebra</td>\n",
+       "      <td>derecho</td>\n",
+       "      <td>humanitario</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>10</th>\n",
+       "      <td>10</td>\n",
+       "      <td>0.083349</td>\n",
+       "      <td>el</td>\n",
+       "      <td>convenio</td>\n",
+       "      <td>de</td>\n",
+       "      <td>ginebra</td>\n",
+       "      <td>derecho</td>\n",
+       "      <td>humanitario</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>11</th>\n",
+       "      <td>11</td>\n",
+       "      <td>0.081911</td>\n",
+       "      <td>el</td>\n",
+       "      <td>convenio</td>\n",
+       "      <td>de</td>\n",
+       "      <td>ginebra</td>\n",
+       "      <td>derecho</td>\n",
+       "      <td>humanitario</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "    rule    weight word1     word2 word3    word4    word5        word6\n",
+       "0      0  0.081911    el  convenio    de  ginebra  derecho  humanitario\n",
+       "1      1  0.087933    el  convenio    de  ginebra  derecho  humanitario\n",
+       "2      2  0.085374    el  convenio    de  ginebra  derecho  humanitario\n",
+       "3      3  0.083864    el  convenio    de  ginebra  derecho  humanitario\n",
+       "4      4  0.083349    el  convenio    de  ginebra  derecho  humanitario\n",
+       "5      5  0.081911    el  convenio    de  ginebra  derecho  humanitario\n",
+       "6      6  0.082985    el  convenio    de  ginebra  derecho  humanitario\n",
+       "7      7  0.082480    el  convenio    de  ginebra  derecho  humanitario\n",
+       "8      8  0.081071    el  convenio    de  ginebra  derecho  humanitario\n",
+       "9      9  0.083864    el  convenio    de  ginebra  derecho  humanitario\n",
+       "10    10  0.083349    el  convenio    de  ginebra  derecho  humanitario\n",
+       "11    11  0.081911    el  convenio    de  ginebra  derecho  humanitario"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.04779147271034485\n",
+      "model : LinearSVM , score = 0.058570542697414935\n",
+      "model : RBFSVM , score = 0.0854309586301861\n",
+      "model : DecisionTree , score = 0.02643110905790252\n",
+      "model : RandomForest , score = 0.008685935557837875\n",
+      "model : AdaBoost , score = 0.019361885843466525\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 27+21_106_29+10_30_106_29+10_100_29_106_29+11_30_106_29+11_100_29_106_29+33_1_30_106_29+33_1_100_29_106_29+.csv\n",
+      "Rules(classes) number : 8\n",
+      "Words(features) number : 6\n",
+      "Records number : 264"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "      <th>word5</th>\n",
+       "      <th>word6</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.127872</td>\n",
+       "      <td>uno</td>\n",
+       "      <td>papel</td>\n",
+       "      <td>más</td>\n",
+       "      <td>activo</td>\n",
+       "      <td>y</td>\n",
+       "      <td>independiente</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.121875</td>\n",
+       "      <td>uno</td>\n",
+       "      <td>papel</td>\n",
+       "      <td>más</td>\n",
+       "      <td>activo</td>\n",
+       "      <td>y</td>\n",
+       "      <td>independiente</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.124957</td>\n",
+       "      <td>uno</td>\n",
+       "      <td>papel</td>\n",
+       "      <td>más</td>\n",
+       "      <td>activo</td>\n",
+       "      <td>y</td>\n",
+       "      <td>independiente</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>0.124957</td>\n",
+       "      <td>uno</td>\n",
+       "      <td>papel</td>\n",
+       "      <td>más</td>\n",
+       "      <td>activo</td>\n",
+       "      <td>y</td>\n",
+       "      <td>independiente</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>4</td>\n",
+       "      <td>0.125213</td>\n",
+       "      <td>uno</td>\n",
+       "      <td>papel</td>\n",
+       "      <td>más</td>\n",
+       "      <td>activo</td>\n",
+       "      <td>y</td>\n",
+       "      <td>independiente</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>5</th>\n",
+       "      <td>5</td>\n",
+       "      <td>0.125213</td>\n",
+       "      <td>uno</td>\n",
+       "      <td>papel</td>\n",
+       "      <td>más</td>\n",
+       "      <td>activo</td>\n",
+       "      <td>y</td>\n",
+       "      <td>independiente</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>6</th>\n",
+       "      <td>6</td>\n",
+       "      <td>0.124957</td>\n",
+       "      <td>uno</td>\n",
+       "      <td>papel</td>\n",
+       "      <td>más</td>\n",
+       "      <td>activo</td>\n",
+       "      <td>y</td>\n",
+       "      <td>independiente</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7</th>\n",
+       "      <td>7</td>\n",
+       "      <td>0.124957</td>\n",
+       "      <td>uno</td>\n",
+       "      <td>papel</td>\n",
+       "      <td>más</td>\n",
+       "      <td>activo</td>\n",
+       "      <td>y</td>\n",
+       "      <td>independiente</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight word1  word2 word3   word4 word5          word6\n",
+       "0     0  0.127872   uno  papel   más  activo     y  independiente\n",
+       "1     1  0.121875   uno  papel   más  activo     y  independiente\n",
+       "2     2  0.124957   uno  papel   más  activo     y  independiente\n",
+       "3     3  0.124957   uno  papel   más  activo     y  independiente\n",
+       "4     4  0.125213   uno  papel   más  activo     y  independiente\n",
+       "5     5  0.125213   uno  papel   más  activo     y  independiente\n",
+       "6     6  0.124957   uno  papel   más  activo     y  independiente\n",
+       "7     7  0.124957   uno  papel   más  activo     y  independiente"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.0542559900037661\n",
+      "model : LinearSVM , score = 0.08926087280201091\n",
+      "model : RBFSVM , score = 0.12326377231319587\n",
+      "model : DecisionTree , score = 0.0\n",
+      "model : RandomForest , score = 0.0\n",
+      "model : AdaBoost , score = 0.0769822895774275\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 20_106_29+12_29_106_29+113_23+113_24+113_14_106_29+113_1_29_106_29+.csv\n",
+      "Rules(classes) number : 6\n",
+      "Words(features) number : 5\n",
+      "Records number : 12"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "      <th>word5</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.168775</td>\n",
+       "      <td>seis</td>\n",
+       "      <td>objetivo</td>\n",
+       "      <td>diferente</td>\n",
+       "      <td>y</td>\n",
+       "      <td>diverso</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.165523</td>\n",
+       "      <td>seis</td>\n",
+       "      <td>objetivo</td>\n",
+       "      <td>diferente</td>\n",
+       "      <td>y</td>\n",
+       "      <td>diverso</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.162509</td>\n",
+       "      <td>seis</td>\n",
+       "      <td>objetivo</td>\n",
+       "      <td>diferente</td>\n",
+       "      <td>y</td>\n",
+       "      <td>diverso</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>0.168894</td>\n",
+       "      <td>seis</td>\n",
+       "      <td>objetivo</td>\n",
+       "      <td>diferente</td>\n",
+       "      <td>y</td>\n",
+       "      <td>diverso</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>4</td>\n",
+       "      <td>0.168775</td>\n",
+       "      <td>seis</td>\n",
+       "      <td>objetivo</td>\n",
+       "      <td>diferente</td>\n",
+       "      <td>y</td>\n",
+       "      <td>diverso</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>5</th>\n",
+       "      <td>5</td>\n",
+       "      <td>0.165523</td>\n",
+       "      <td>seis</td>\n",
+       "      <td>objetivo</td>\n",
+       "      <td>diferente</td>\n",
+       "      <td>y</td>\n",
+       "      <td>diverso</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight word1     word2      word3 word4    word5\n",
+       "0     0  0.168775  seis  objetivo  diferente     y  diverso\n",
+       "1     1  0.165523  seis  objetivo  diferente     y  diverso\n",
+       "2     2  0.162509  seis  objetivo  diferente     y  diverso\n",
+       "3     3  0.168894  seis  objetivo  diferente     y  diverso\n",
+       "4     4  0.168775  seis  objetivo  diferente     y  diverso\n",
+       "5     5  0.165523  seis  objetivo  diferente     y  diverso"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.1687751687751688\n",
+      "model : LinearSVM , score = 0.16889416889416892\n",
+      "model : RBFSVM , score = 0.16889416889416892\n",
+      "model : DecisionTree , score = 0.16889416889416892\n",
+      "model : RandomForest , score = 0.1687751687751688\n",
+      "model : AdaBoost , score = 0.16889416889416892\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 24+1_63_106_29+.csv\n",
+      "Rules(classes) number : 2\n",
+      "Words(features) number : 4\n",
+      "Records number : 6"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "/home/aboelhamd/anaconda3/lib/python3.7/site-packages/sklearn/naive_bayes.py:436: RuntimeWarning: divide by zero encountered in log\n",
+      "  n_ij = - 0.5 * np.sum(np.log(2. * np.pi * self.sigma_[i, :]))\n",
+      "/home/aboelhamd/anaconda3/lib/python3.7/site-packages/sklearn/naive_bayes.py:438: RuntimeWarning: invalid value encountered in true_divide\n",
+      "  (self.sigma_[i, :]), 1)\n"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.505526</td>\n",
+       "      <td>inmigración</td>\n",
+       "      <td>equilibrar</td>\n",
+       "      <td>y</td>\n",
+       "      <td>coherente</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.494474</td>\n",
+       "      <td>inmigración</td>\n",
+       "      <td>equilibrar</td>\n",
+       "      <td>y</td>\n",
+       "      <td>coherente</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight        word1       word2 word3      word4\n",
+       "0     0  0.505526  inmigración  equilibrar     y  coherente\n",
+       "1     1  0.494474  inmigración  equilibrar     y  coherente"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.3298335598322446\n",
+      "model : LinearSVM , score = 0.3298335598322446\n",
+      "model : RBFSVM , score = 0.3298335598322446\n",
+      "model : DecisionTree , score = 0.3298335598322446\n",
+      "model : RandomForest , score = 0.3298335598322446\n",
+      "model : AdaBoost , score = 0.3298335598322446\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 96+99_108_113+113_108_1_108_113+.csv\n",
+      "Rules(classes) number : 3\n",
+      "Words(features) number : 5\n",
+      "Records number : 255"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "      <th>word5</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.376928</td>\n",
+       "      <td>1</td>\n",
+       "      <td>de</td>\n",
+       "      <td>enero</td>\n",
+       "      <td>de</td>\n",
+       "      <td>2003</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.329794</td>\n",
+       "      <td>1</td>\n",
+       "      <td>de</td>\n",
+       "      <td>enero</td>\n",
+       "      <td>de</td>\n",
+       "      <td>2003</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.293278</td>\n",
+       "      <td>1</td>\n",
+       "      <td>de</td>\n",
+       "      <td>enero</td>\n",
+       "      <td>de</td>\n",
+       "      <td>2003</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight  word1 word2  word3 word4  word5\n",
+       "0     0  0.376928      1    de  enero    de   2003\n",
+       "1     1  0.329794      1    de  enero    de   2003\n",
+       "2     2  0.293278      1    de  enero    de   2003"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.28492173347494804\n",
+      "model : LinearSVM , score = 0.3367974223858633\n",
+      "model : RBFSVM , score = 0.39453017910222304\n",
+      "model : DecisionTree , score = 0.17837654888961102\n",
+      "model : RandomForest , score = 0.1659894936543952\n",
+      "model : AdaBoost , score = 0.2195843182352461\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 30+100_63+.csv\n",
+      "Rules(classes) number : 2\n",
+      "Words(features) number : 2\n",
+      "Records number : 1076"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.5</td>\n",
+       "      <td>de_alguna_manera</td>\n",
+       "      <td>ocultar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.5</td>\n",
+       "      <td>de_alguna_manera</td>\n",
+       "      <td>ocultar</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule  weight             word1    word2\n",
+       "0     0     0.5  de_alguna_manera  ocultar\n",
+       "1     1     0.5  de_alguna_manera  ocultar"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.4572205314900661\n",
+      "model : LinearSVM , score = 0.4647990828862173\n",
+      "model : RBFSVM , score = 0.2732230548762148\n",
+      "model : DecisionTree , score = 0.4237193577408349\n",
+      "model : RandomForest , score = 0.38300158878767293\n",
+      "model : AdaBoost , score = 0.3903416770705051\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 84_69+100_73+100_47_69+.csv\n",
+      "Rules(classes) number : 3\n",
+      "Words(features) number : 4\n",
+      "Records number : 15"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.296842</td>\n",
+       "      <td>hoy</td>\n",
+       "      <td>lo</td>\n",
+       "      <td>estar</td>\n",
+       "      <td>debatir</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.406315</td>\n",
+       "      <td>hoy</td>\n",
+       "      <td>lo</td>\n",
+       "      <td>estar</td>\n",
+       "      <td>debatir</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.296842</td>\n",
+       "      <td>hoy</td>\n",
+       "      <td>lo</td>\n",
+       "      <td>estar</td>\n",
+       "      <td>debatir</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight word1 word2  word3    word4\n",
+       "0     0  0.296842   hoy    lo  estar  debatir\n",
+       "1     1  0.406315   hoy    lo  estar  debatir\n",
+       "2     2  0.296842   hoy    lo  estar  debatir"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.11115762827378055\n",
+      "model : LinearSVM , score = 0.2336983141481243\n",
+      "model : RBFSVM , score = 0.2336983141481243\n",
+      "model : DecisionTree , score = 0.11115762827378055\n",
+      "model : RandomForest , score = 0.11115762827378055\n",
+      "model : AdaBoost , score = 0.11115762827378055\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 10_63_29+11_63_29+33_15+33_1_63_29+.csv\n",
+      "Rules(classes) number : 4\n",
+      "Words(features) number : 4\n",
+      "Records number : 84"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.247563</td>\n",
+       "      <td>uno</td>\n",
+       "      <td>sistema</td>\n",
+       "      <td>informatizar</td>\n",
+       "      <td>eficaz</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.247210</td>\n",
+       "      <td>uno</td>\n",
+       "      <td>sistema</td>\n",
+       "      <td>informatizar</td>\n",
+       "      <td>eficaz</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.257664</td>\n",
+       "      <td>uno</td>\n",
+       "      <td>sistema</td>\n",
+       "      <td>informatizar</td>\n",
+       "      <td>eficaz</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>0.247563</td>\n",
+       "      <td>uno</td>\n",
+       "      <td>sistema</td>\n",
+       "      <td>informatizar</td>\n",
+       "      <td>eficaz</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight word1    word2         word3   word4\n",
+       "0     0  0.247563   uno  sistema  informatizar  eficaz\n",
+       "1     1  0.247210   uno  sistema  informatizar  eficaz\n",
+       "2     2  0.257664   uno  sistema  informatizar  eficaz\n",
+       "3     3  0.247563   uno  sistema  informatizar  eficaz"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.11849355553125353\n",
+      "model : LinearSVM , score = 0.19053970371076076\n",
+      "model : RBFSVM , score = 0.23823980418499188\n",
+      "model : DecisionTree , score = 0.0694058122852295\n",
+      "model : RandomForest , score = 0.046491304325740085\n",
+      "model : AdaBoost , score = 0.11843892392606487\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 64_64+64_108_68+108_38+108_68_64+108_68_108_68+.csv\n",
+      "Rules(classes) number : 5\n",
+      "Words(features) number : 4\n",
+      "Records number : 25"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.202745</td>\n",
+       "      <td>para</td>\n",
+       "      <td>llegar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>conseguir</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.202745</td>\n",
+       "      <td>para</td>\n",
+       "      <td>llegar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>conseguir</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.204575</td>\n",
+       "      <td>para</td>\n",
+       "      <td>llegar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>conseguir</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>0.194967</td>\n",
+       "      <td>para</td>\n",
+       "      <td>llegar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>conseguir</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>4</td>\n",
+       "      <td>0.194967</td>\n",
+       "      <td>para</td>\n",
+       "      <td>llegar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>conseguir</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight word1   word2 word3      word4\n",
+       "0     0  0.202745  para  llegar     a  conseguir\n",
+       "1     1  0.202745  para  llegar     a  conseguir\n",
+       "2     2  0.204575  para  llegar     a  conseguir\n",
+       "3     3  0.194967  para  llegar     a  conseguir\n",
+       "4     4  0.194967  para  llegar     a  conseguir"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.1535092359942748\n",
+      "model : LinearSVM , score = 0.15974950555717687\n",
+      "model : RBFSVM , score = 0.15974950555717687\n",
+      "model : DecisionTree , score = 0.07794581534522631\n",
+      "model : RandomForest , score = 0.07560799123470986\n",
+      "model : AdaBoost , score = 0.15974950555717687\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 60+47_68+.csv\n",
+      "Rules(classes) number : 2\n",
+      "Words(features) number : 3\n",
+      "Records number : 16"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.5</td>\n",
+       "      <td>lo</td>\n",
+       "      <td>poder</td>\n",
+       "      <td>constatar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.5</td>\n",
+       "      <td>lo</td>\n",
+       "      <td>poder</td>\n",
+       "      <td>constatar</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule  weight word1  word2      word3\n",
+       "0     0     0.5    lo  poder  constatar\n",
+       "1     1     0.5    lo  poder  constatar"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.3907131479265588\n",
+      "model : LinearSVM , score = 0.5187070811770713\n",
+      "model : RBFSVM , score = 0.5187070811770713\n",
+      "model : DecisionTree , score = 0.3907131479265588\n",
+      "model : RandomForest , score = 0.3907131479265588\n",
+      "model : AdaBoost , score = 0.3907131479265588\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 38+37_64+37_108_68+.csv\n",
+      "Rules(classes) number : 3\n",
+      "Words(features) number : 3\n",
+      "Records number : 51"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.354531</td>\n",
+       "      <td>llegar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>ser</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.322735</td>\n",
+       "      <td>llegar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>ser</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.322735</td>\n",
+       "      <td>llegar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>ser</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight   word1 word2 word3\n",
+       "0     0  0.354531  llegar     a   ser\n",
+       "1     1  0.322735  llegar     a   ser\n",
+       "2     2  0.322735  llegar     a   ser"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.23554661057521825\n",
+      "model : LinearSVM , score = 0.2961067303875821\n",
+      "model : RBFSVM , score = 0.18448152174300372\n",
+      "model : DecisionTree , score = 0.1981366276139268\n",
+      "model : RandomForest , score = 0.19869171041483163\n",
+      "model : AdaBoost , score = 0.1981366276139268\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 32+33_30+33_100_29+.csv\n",
+      "Rules(classes) number : 3\n",
+      "Words(features) number : 3\n",
+      "Records number : 456"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.340749</td>\n",
+       "      <td>el</td>\n",
+       "      <td>más</td>\n",
+       "      <td>moderno</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.329626</td>\n",
+       "      <td>el</td>\n",
+       "      <td>más</td>\n",
+       "      <td>moderno</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.329626</td>\n",
+       "      <td>el</td>\n",
+       "      <td>más</td>\n",
+       "      <td>moderno</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight word1 word2    word3\n",
+       "0     0  0.340749    el   más  moderno\n",
+       "1     1  0.329626    el   más  moderno\n",
+       "2     2  0.329626    el   más  moderno"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.31015551879539377\n",
+      "model : LinearSVM , score = 0.3302076467085666\n",
+      "model : RBFSVM , score = 0.33077810202475727\n",
+      "model : DecisionTree , score = 0.27999006952059113\n",
+      "model : RandomForest , score = 0.2324501225530451\n",
+      "model : AdaBoost , score = 0.23169006359161295\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 40+69_63+.csv\n",
+      "Rules(classes) number : 2\n",
+      "Words(features) number : 2\n",
+      "Records number : 8"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.5</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>escuchar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.5</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>escuchar</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule  weight  word1     word2\n",
+       "0     0     0.5  haber  escuchar\n",
+       "1     1     0.5  haber  escuchar"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.25\n",
+      "model : LinearSVM , score = 0.25\n",
+      "model : RBFSVM , score = 0.25\n",
+      "model : DecisionTree , score = 0.25\n",
+      "model : RandomForest , score = 0.25\n",
+      "model : AdaBoost , score = 0.25\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 18_106_63+17_29_106_63+31_23+31_14_106_63+31_1_29_106_63+33_29_23+33_29_14_106_63+33_29_1_29_106_63+.csv\n",
+      "Rules(classes) number : 8\n",
+      "Words(features) number : 6\n",
+      "Records number : 16"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "      <th>word5</th>\n",
+       "      <th>word6</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.130386</td>\n",
+       "      <td>el</td>\n",
+       "      <td>actual</td>\n",
+       "      <td>estructura</td>\n",
+       "      <td>presupuestario</td>\n",
+       "      <td>y</td>\n",
+       "      <td>determinar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.121098</td>\n",
+       "      <td>el</td>\n",
+       "      <td>actual</td>\n",
+       "      <td>estructura</td>\n",
+       "      <td>presupuestario</td>\n",
+       "      <td>y</td>\n",
+       "      <td>determinar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.122773</td>\n",
+       "      <td>el</td>\n",
+       "      <td>actual</td>\n",
+       "      <td>estructura</td>\n",
+       "      <td>presupuestario</td>\n",
+       "      <td>y</td>\n",
+       "      <td>determinar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>0.130386</td>\n",
+       "      <td>el</td>\n",
+       "      <td>actual</td>\n",
+       "      <td>estructura</td>\n",
+       "      <td>presupuestario</td>\n",
+       "      <td>y</td>\n",
+       "      <td>determinar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>4</td>\n",
+       "      <td>0.121098</td>\n",
+       "      <td>el</td>\n",
+       "      <td>actual</td>\n",
+       "      <td>estructura</td>\n",
+       "      <td>presupuestario</td>\n",
+       "      <td>y</td>\n",
+       "      <td>determinar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>5</th>\n",
+       "      <td>5</td>\n",
+       "      <td>0.122773</td>\n",
+       "      <td>el</td>\n",
+       "      <td>actual</td>\n",
+       "      <td>estructura</td>\n",
+       "      <td>presupuestario</td>\n",
+       "      <td>y</td>\n",
+       "      <td>determinar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>6</th>\n",
+       "      <td>6</td>\n",
+       "      <td>0.130386</td>\n",
+       "      <td>el</td>\n",
+       "      <td>actual</td>\n",
+       "      <td>estructura</td>\n",
+       "      <td>presupuestario</td>\n",
+       "      <td>y</td>\n",
+       "      <td>determinar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7</th>\n",
+       "      <td>7</td>\n",
+       "      <td>0.121098</td>\n",
+       "      <td>el</td>\n",
+       "      <td>actual</td>\n",
+       "      <td>estructura</td>\n",
+       "      <td>presupuestario</td>\n",
+       "      <td>y</td>\n",
+       "      <td>determinar</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight word1   word2       word3           word4 word5       word6\n",
+       "0     0  0.130386    el  actual  estructura  presupuestario     y  determinar\n",
+       "1     1  0.121098    el  actual  estructura  presupuestario     y  determinar\n",
+       "2     2  0.122773    el  actual  estructura  presupuestario     y  determinar\n",
+       "3     3  0.130386    el  actual  estructura  presupuestario     y  determinar\n",
+       "4     4  0.121098    el  actual  estructura  presupuestario     y  determinar\n",
+       "5     5  0.122773    el  actual  estructura  presupuestario     y  determinar\n",
+       "6     6  0.130386    el  actual  estructura  presupuestario     y  determinar\n",
+       "7     7  0.121098    el  actual  estructura  presupuestario     y  determinar"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.13038626077252155\n",
+      "model : LinearSVM , score = 0.13038626077252155\n",
+      "model : RBFSVM , score = 0.13038626077252155\n",
+      "model : DecisionTree , score = 0.13038626077252155\n",
+      "model : RandomForest , score = 0.13038626077252155\n",
+      "model : AdaBoost , score = 0.13038626077252155\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 75+100_36+100_37+.csv\n",
+      "Rules(classes) number : 3\n",
+      "Words(features) number : 2\n",
+      "Records number : 312"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "/home/aboelhamd/anaconda3/lib/python3.7/site-packages/sklearn/naive_bayes.py:436: RuntimeWarning: divide by zero encountered in log\n",
+      "  n_ij = - 0.5 * np.sum(np.log(2. * np.pi * self.sigma_[i, :]))\n",
+      "/home/aboelhamd/anaconda3/lib/python3.7/site-packages/sklearn/naive_bayes.py:438: RuntimeWarning: invalid value encountered in true_divide\n",
+      "  (self.sigma_[i, :]), 1)\n"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.328514</td>\n",
+       "      <td>no</td>\n",
+       "      <td>haber_que</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.335654</td>\n",
+       "      <td>no</td>\n",
+       "      <td>haber_que</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.335831</td>\n",
+       "      <td>no</td>\n",
+       "      <td>haber_que</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight word1      word2\n",
+       "0     0  0.328514    no  haber_que\n",
+       "1     1  0.335654    no  haber_que\n",
+       "2     2  0.335831    no  haber_que"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.297156053940856\n",
+      "model : LinearSVM , score = 0.30364650810851324\n",
+      "model : RBFSVM , score = 0.27833199568512473\n",
+      "model : DecisionTree , score = 0.29661452526235416\n",
+      "model : RandomForest , score = 0.27751982744798837\n",
+      "model : AdaBoost , score = 0.25890578456830055\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 31+33_29+.csv\n",
+      "Rules(classes) number : 2\n",
+      "Words(features) number : 2\n",
+      "Records number : 3450"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.5</td>\n",
+       "      <td>el</td>\n",
+       "      <td>último</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.5</td>\n",
+       "      <td>el</td>\n",
+       "      <td>último</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule  weight word1   word2\n",
+       "0     0     0.5    el  último\n",
+       "1     1     0.5    el  último"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.489499313017107\n",
+      "model : LinearSVM , score = 0.500640208963022\n",
+      "model : RBFSVM , score = 0.40714635803619653\n",
+      "model : DecisionTree , score = 0.4665642676145503\n",
+      "model : RandomForest , score = 0.45032754557246746\n",
+      "model : AdaBoost , score = 0.4576052390901414\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 39+40_64+40_108_68+37_38+37_67+37_63_64+37_63_108_68+.csv\n",
+      "Rules(classes) number : 7\n",
+      "Words(features) number : 4\n",
+      "Records number : 28"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.159656</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>llegar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>alcanzar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.136138</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>llegar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>alcanzar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.136138</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>llegar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>alcanzar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>0.159656</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>llegar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>alcanzar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>4</td>\n",
+       "      <td>0.136138</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>llegar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>alcanzar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>5</th>\n",
+       "      <td>5</td>\n",
+       "      <td>0.136138</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>llegar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>alcanzar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>6</th>\n",
+       "      <td>6</td>\n",
+       "      <td>0.136138</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>llegar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>alcanzar</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight  word1   word2 word3     word4\n",
+       "0     0  0.159656  haber  llegar     a  alcanzar\n",
+       "1     1  0.136138  haber  llegar     a  alcanzar\n",
+       "2     2  0.136138  haber  llegar     a  alcanzar\n",
+       "3     3  0.159656  haber  llegar     a  alcanzar\n",
+       "4     4  0.136138  haber  llegar     a  alcanzar\n",
+       "5     5  0.136138  haber  llegar     a  alcanzar\n",
+       "6     6  0.136138  haber  llegar     a  alcanzar"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.0\n",
+      "model : LinearSVM , score = 0.15303900968914755\n",
+      "model : RBFSVM , score = 0.15303900968914755\n",
+      "model : DecisionTree , score = 0.0\n",
+      "model : RandomForest , score = 0.0\n",
+      "model : AdaBoost , score = 0.0\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 32_64+32_108_68+33_30_64+33_30_108_68+33_100_67+33_100_63_64+33_100_63_108_68+.csv\n",
+      "Rules(classes) number : 7\n",
+      "Words(features) number : 5\n",
+      "Records number : 14"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "      <th>word5</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.148288</td>\n",
+       "      <td>el</td>\n",
+       "      <td>más</td>\n",
+       "      <td>cualificar</td>\n",
+       "      <td>para</td>\n",
+       "      <td>organizar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.140989</td>\n",
+       "      <td>el</td>\n",
+       "      <td>más</td>\n",
+       "      <td>cualificar</td>\n",
+       "      <td>para</td>\n",
+       "      <td>organizar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.145447</td>\n",
+       "      <td>el</td>\n",
+       "      <td>más</td>\n",
+       "      <td>cualificar</td>\n",
+       "      <td>para</td>\n",
+       "      <td>organizar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>0.137192</td>\n",
+       "      <td>el</td>\n",
+       "      <td>más</td>\n",
+       "      <td>cualificar</td>\n",
+       "      <td>para</td>\n",
+       "      <td>organizar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>4</td>\n",
+       "      <td>0.145447</td>\n",
+       "      <td>el</td>\n",
+       "      <td>más</td>\n",
+       "      <td>cualificar</td>\n",
+       "      <td>para</td>\n",
+       "      <td>organizar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>5</th>\n",
+       "      <td>5</td>\n",
+       "      <td>0.145447</td>\n",
+       "      <td>el</td>\n",
+       "      <td>más</td>\n",
+       "      <td>cualificar</td>\n",
+       "      <td>para</td>\n",
+       "      <td>organizar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>6</th>\n",
+       "      <td>6</td>\n",
+       "      <td>0.137192</td>\n",
+       "      <td>el</td>\n",
+       "      <td>más</td>\n",
+       "      <td>cualificar</td>\n",
+       "      <td>para</td>\n",
+       "      <td>organizar</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight word1 word2       word3 word4      word5\n",
+       "0     0  0.148288    el   más  cualificar  para  organizar\n",
+       "1     1  0.140989    el   más  cualificar  para  organizar\n",
+       "2     2  0.145447    el   más  cualificar  para  organizar\n",
+       "3     3  0.137192    el   más  cualificar  para  organizar\n",
+       "4     4  0.145447    el   más  cualificar  para  organizar\n",
+       "5     5  0.145447    el   más  cualificar  para  organizar\n",
+       "6     6  0.137192    el   más  cualificar  para  organizar"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.14828770342459316\n",
+      "model : LinearSVM , score = 0.14828770342459316\n",
+      "model : RBFSVM , score = 0.14828770342459316\n",
+      "model : DecisionTree , score = 0.14828770342459316\n",
+      "model : RandomForest , score = 0.14828770342459316\n",
+      "model : AdaBoost , score = 0.14828770342459316\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 76+100_40+100_69_63+.csv\n",
+      "Rules(classes) number : 3\n",
+      "Words(features) number : 3\n",
+      "Records number : 6"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "/home/aboelhamd/anaconda3/lib/python3.7/site-packages/sklearn/naive_bayes.py:436: RuntimeWarning: divide by zero encountered in log\n",
+      "  n_ij = - 0.5 * np.sum(np.log(2. * np.pi * self.sigma_[i, :]))\n",
+      "/home/aboelhamd/anaconda3/lib/python3.7/site-packages/sklearn/naive_bayes.py:438: RuntimeWarning: invalid value encountered in true_divide\n",
+      "  (self.sigma_[i, :]), 1)\n"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.335418</td>\n",
+       "      <td>no</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>presentar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.317747</td>\n",
+       "      <td>no</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>presentar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.346834</td>\n",
+       "      <td>no</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>presentar</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight word1  word2      word3\n",
+       "0     0  0.335418    no  haber  presentar\n",
+       "1     1  0.317747    no  haber  presentar\n",
+       "2     2  0.346834    no  haber  presentar"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.3354183354183354\n",
+      "model : LinearSVM , score = 0.34683434683434683\n",
+      "model : RBFSVM , score = 0.34683434683434683\n",
+      "model : DecisionTree , score = 0.34683434683434683\n",
+      "model : RandomForest , score = 0.31774731774731774\n",
+      "model : AdaBoost ,"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "/home/aboelhamd/anaconda3/lib/python3.7/site-packages/sklearn/naive_bayes.py:436: RuntimeWarning: divide by zero encountered in log\n",
+      "  n_ij = - 0.5 * np.sum(np.log(2. * np.pi * self.sigma_[i, :]))\n",
+      "/home/aboelhamd/anaconda3/lib/python3.7/site-packages/sklearn/naive_bayes.py:438: RuntimeWarning: invalid value encountered in true_divide\n",
+      "  (self.sigma_[i, :]), 1)\n"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      " score = 0.34683434683434683\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 40_64+40_108_68+68_67+68_63_64+68_63_108_68+.csv\n",
+      "Rules(classes) number : 5\n",
+      "Words(features) number : 4\n",
+      "Records number : 10"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.2</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>limitar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>realizar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.2</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>limitar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>realizar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.2</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>limitar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>realizar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>0.2</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>limitar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>realizar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>4</td>\n",
+       "      <td>0.2</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>limitar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>realizar</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule  weight  word1    word2 word3     word4\n",
+       "0     0     0.2  haber  limitar     a  realizar\n",
+       "1     1     0.2  haber  limitar     a  realizar\n",
+       "2     2     0.2  haber  limitar     a  realizar\n",
+       "3     3     0.2  haber  limitar     a  realizar\n",
+       "4     4     0.2  haber  limitar     a  realizar"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.0\n",
+      "model : LinearSVM , score = 0.20244410771241195\n",
+      "model : RBFSVM , score = 0.20244410771241195\n",
+      "model : DecisionTree , score = 0.0\n",
+      "model : RandomForest , score = 0.0\n",
+      "model : AdaBoost , score = 0.0\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 99_29+113_108_14+113_108_1_29+.csv\n",
+      "Rules(classes) number : 3\n",
+      "Words(features) number : 4\n",
+      "Records number : 6"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.336890</td>\n",
+       "      <td>21</td>\n",
+       "      <td>de</td>\n",
+       "      <td>septiembre</td>\n",
+       "      <td>próximo</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.337982</td>\n",
+       "      <td>21</td>\n",
+       "      <td>de</td>\n",
+       "      <td>septiembre</td>\n",
+       "      <td>próximo</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.325129</td>\n",
+       "      <td>21</td>\n",
+       "      <td>de</td>\n",
+       "      <td>septiembre</td>\n",
+       "      <td>próximo</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight  word1 word2       word3    word4\n",
+       "0     0  0.336890     21    de  septiembre  próximo\n",
+       "1     1  0.337982     21    de  septiembre  próximo\n",
+       "2     2  0.325129     21    de  septiembre  próximo"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.33688966311033686\n",
+      "model : LinearSVM , score = 0.33798166201833796\n",
+      "model : RBFSVM , score = 0.33798166201833796\n",
+      "model : DecisionTree , score = 0.33798166201833796\n",
+      "model : RandomForest , score = 0.33688966311033686\n",
+      "model : AdaBoost , score = 0.33798166201833796\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 22+12_30+12_110_63+113_1_30+113_1_110_63+.csv\n",
+      "Rules(classes) number : 5\n",
+      "Words(features) number : 4\n",
+      "Records number : 10"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "/home/aboelhamd/anaconda3/lib/python3.7/site-packages/sklearn/naive_bayes.py:436: RuntimeWarning: divide by zero encountered in log\n",
+      "  n_ij = - 0.5 * np.sum(np.log(2. * np.pi * self.sigma_[i, :]))\n",
+      "/home/aboelhamd/anaconda3/lib/python3.7/site-packages/sklearn/naive_bayes.py:438: RuntimeWarning: invalid value encountered in true_divide\n",
+      "  (self.sigma_[i, :]), 1)\n"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.208152</td>\n",
+       "      <td>dos</td>\n",
+       "      <td>colega</td>\n",
+       "      <td>tan</td>\n",
+       "      <td>experimentar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.197962</td>\n",
+       "      <td>dos</td>\n",
+       "      <td>colega</td>\n",
+       "      <td>tan</td>\n",
+       "      <td>experimentar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.197962</td>\n",
+       "      <td>dos</td>\n",
+       "      <td>colega</td>\n",
+       "      <td>tan</td>\n",
+       "      <td>experimentar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>0.197962</td>\n",
+       "      <td>dos</td>\n",
+       "      <td>colega</td>\n",
+       "      <td>tan</td>\n",
+       "      <td>experimentar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>4</td>\n",
+       "      <td>0.197962</td>\n",
+       "      <td>dos</td>\n",
+       "      <td>colega</td>\n",
+       "      <td>tan</td>\n",
+       "      <td>experimentar</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight word1   word2 word3         word4\n",
+       "0     0  0.208152   dos  colega   tan  experimentar\n",
+       "1     1  0.197962   dos  colega   tan  experimentar\n",
+       "2     2  0.197962   dos  colega   tan  experimentar\n",
+       "3     3  0.197962   dos  colega   tan  experimentar\n",
+       "4     4  0.197962   dos  colega   tan  experimentar"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.208152\n",
+      "model : LinearSVM , score = 0.208152\n",
+      "model : RBFSVM , score = 0.208152\n",
+      "model : DecisionTree , score = 0.208152\n",
+      "model : RandomForest , score = 0.208152\n",
+      "model : AdaBoost ,"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "/home/aboelhamd/anaconda3/lib/python3.7/site-packages/sklearn/naive_bayes.py:436: RuntimeWarning: divide by zero encountered in log\n",
+      "  n_ij = - 0.5 * np.sum(np.log(2. * np.pi * self.sigma_[i, :]))\n",
+      "/home/aboelhamd/anaconda3/lib/python3.7/site-packages/sklearn/naive_bayes.py:438: RuntimeWarning: invalid value encountered in true_divide\n",
+      "  (self.sigma_[i, :]), 1)\n"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      " score = 0.208152\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 36+37+.csv\n",
+      "Rules(classes) number : 2\n",
+      "Words(features) number : 1\n",
+      "Records number : 836"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.506138</td>\n",
+       "      <td>haber_que</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.493862</td>\n",
+       "      <td>haber_que</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight      word1\n",
+       "0     0  0.506138  haber_que\n",
+       "1     1  0.493862  haber_que"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.5010786051719591\n",
+      "model : LinearSVM , score = 0.5010786051719591\n",
+      "model : RBFSVM , score = 0.5010786051719591\n",
+      "model : DecisionTree , score = 0.5010786051719591\n",
+      "model : RandomForest , score = 0.5010786051719591\n",
+      "model : AdaBoost ,"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "/home/aboelhamd/anaconda3/lib/python3.7/site-packages/sklearn/naive_bayes.py:436: RuntimeWarning: divide by zero encountered in log\n",
+      "  n_ij = - 0.5 * np.sum(np.log(2. * np.pi * self.sigma_[i, :]))\n",
+      "/home/aboelhamd/anaconda3/lib/python3.7/site-packages/sklearn/naive_bayes.py:438: RuntimeWarning: invalid value encountered in true_divide\n",
+      "  (self.sigma_[i, :]), 1)\n"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      " score = 0.5010786051719591\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 18_106+17_29_106+31_24+31_14_106+31_1_29_106+33_29_24+33_29_14_106+33_29_1_29_106+.csv\n",
+      "Rules(classes) number : 8\n",
+      "Words(features) number : 5\n",
+      "Records number : 216"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "      <th>word5</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.125313</td>\n",
+       "      <td>el</td>\n",
+       "      <td>distinto</td>\n",
+       "      <td>condición</td>\n",
+       "      <td>geográfico</td>\n",
+       "      <td>y</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.123135</td>\n",
+       "      <td>el</td>\n",
+       "      <td>distinto</td>\n",
+       "      <td>condición</td>\n",
+       "      <td>geográfico</td>\n",
+       "      <td>y</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.127328</td>\n",
+       "      <td>el</td>\n",
+       "      <td>distinto</td>\n",
+       "      <td>condición</td>\n",
+       "      <td>geográfico</td>\n",
+       "      <td>y</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>0.125313</td>\n",
+       "      <td>el</td>\n",
+       "      <td>distinto</td>\n",
+       "      <td>condición</td>\n",
+       "      <td>geográfico</td>\n",
+       "      <td>y</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>4</td>\n",
+       "      <td>0.123135</td>\n",
+       "      <td>el</td>\n",
+       "      <td>distinto</td>\n",
+       "      <td>condición</td>\n",
+       "      <td>geográfico</td>\n",
+       "      <td>y</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>5</th>\n",
+       "      <td>5</td>\n",
+       "      <td>0.127328</td>\n",
+       "      <td>el</td>\n",
+       "      <td>distinto</td>\n",
+       "      <td>condición</td>\n",
+       "      <td>geográfico</td>\n",
+       "      <td>y</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>6</th>\n",
+       "      <td>6</td>\n",
+       "      <td>0.125313</td>\n",
+       "      <td>el</td>\n",
+       "      <td>distinto</td>\n",
+       "      <td>condición</td>\n",
+       "      <td>geográfico</td>\n",
+       "      <td>y</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7</th>\n",
+       "      <td>7</td>\n",
+       "      <td>0.123135</td>\n",
+       "      <td>el</td>\n",
+       "      <td>distinto</td>\n",
+       "      <td>condición</td>\n",
+       "      <td>geográfico</td>\n",
+       "      <td>y</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight word1     word2      word3       word4 word5\n",
+       "0     0  0.125313    el  distinto  condición  geográfico     y\n",
+       "1     1  0.123135    el  distinto  condición  geográfico     y\n",
+       "2     2  0.127328    el  distinto  condición  geográfico     y\n",
+       "3     3  0.125313    el  distinto  condición  geográfico     y\n",
+       "4     4  0.123135    el  distinto  condición  geográfico     y\n",
+       "5     5  0.127328    el  distinto  condición  geográfico     y\n",
+       "6     6  0.125313    el  distinto  condición  geográfico     y\n",
+       "7     7  0.123135    el  distinto  condición  geográfico     y"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.08375096177689977\n",
+      "model : LinearSVM , score = 0.06595154882580584\n",
+      "model : RBFSVM , score = 0.12189498544976621\n",
+      "model : DecisionTree , score = 0.009444793672452916\n",
+      "model : RandomForest , score = 0.0\n",
+      "model : AdaBoost , score = 0.04690395659934957\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 67_63+63_65+63_64_63+63_91_63+63_108_40+63_108_68_63+.csv\n",
+      "Rules(classes) number : 6\n",
+      "Words(features) number : 4\n",
+      "Records number : 12"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.164068</td>\n",
+       "      <td>acusar</td>\n",
+       "      <td>de</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>encargar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.153241</td>\n",
+       "      <td>acusar</td>\n",
+       "      <td>de</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>encargar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.164068</td>\n",
+       "      <td>acusar</td>\n",
+       "      <td>de</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>encargar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>0.190487</td>\n",
+       "      <td>acusar</td>\n",
+       "      <td>de</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>encargar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>4</td>\n",
+       "      <td>0.164068</td>\n",
+       "      <td>acusar</td>\n",
+       "      <td>de</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>encargar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>5</th>\n",
+       "      <td>5</td>\n",
+       "      <td>0.164068</td>\n",
+       "      <td>acusar</td>\n",
+       "      <td>de</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>encargar</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight   word1 word2  word3     word4\n",
+       "0     0  0.164068  acusar    de  haber  encargar\n",
+       "1     1  0.153241  acusar    de  haber  encargar\n",
+       "2     2  0.164068  acusar    de  haber  encargar\n",
+       "3     3  0.190487  acusar    de  haber  encargar\n",
+       "4     4  0.164068  acusar    de  haber  encargar\n",
+       "5     5  0.164068  acusar    de  haber  encargar"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.16406800000000002\n",
+      "model : LinearSVM , score = 0.190487\n",
+      "model : RBFSVM , score = 0.190487\n",
+      "model : DecisionTree , score = 0.190487\n",
+      "model : RandomForest , score = 0.16406800000000002\n",
+      "model : AdaBoost , score = 0.190487\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 18+17_29+31_14+31_1_29+33_29_14+33_29_1_29+.csv\n",
+      "Rules(classes) number : 6\n",
+      "Words(features) number : 4\n",
+      "Records number : 5040"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "/home/aboelhamd/anaconda3/lib/python3.7/site-packages/sklearn/naive_bayes.py:436: RuntimeWarning: divide by zero encountered in log\n",
+      "  n_ij = - 0.5 * np.sum(np.log(2. * np.pi * self.sigma_[i, :]))\n",
+      "/home/aboelhamd/anaconda3/lib/python3.7/site-packages/sklearn/naive_bayes.py:438: RuntimeWarning: invalid value encountered in true_divide\n",
+      "  (self.sigma_[i, :]), 1)\n"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.169627</td>\n",
+       "      <td>suyo</td>\n",
+       "      <td>propio</td>\n",
+       "      <td>análisis</td>\n",
+       "      <td>político</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.163536</td>\n",
+       "      <td>suyo</td>\n",
+       "      <td>propio</td>\n",
+       "      <td>análisis</td>\n",
+       "      <td>político</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.169627</td>\n",
+       "      <td>suyo</td>\n",
+       "      <td>propio</td>\n",
+       "      <td>análisis</td>\n",
+       "      <td>político</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>0.163792</td>\n",
+       "      <td>suyo</td>\n",
+       "      <td>propio</td>\n",
+       "      <td>análisis</td>\n",
+       "      <td>político</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>4</td>\n",
+       "      <td>0.169627</td>\n",
+       "      <td>suyo</td>\n",
+       "      <td>propio</td>\n",
+       "      <td>análisis</td>\n",
+       "      <td>político</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>5</th>\n",
+       "      <td>5</td>\n",
+       "      <td>0.163792</td>\n",
+       "      <td>suyo</td>\n",
+       "      <td>propio</td>\n",
+       "      <td>análisis</td>\n",
+       "      <td>político</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight word1   word2     word3     word4\n",
+       "0     0  0.169627  suyo  propio  análisis  político\n",
+       "1     1  0.163536  suyo  propio  análisis  político\n",
+       "2     2  0.169627  suyo  propio  análisis  político\n",
+       "3     3  0.163792  suyo  propio  análisis  político\n",
+       "4     4  0.169627  suyo  propio  análisis  político\n",
+       "5     5  0.163792  suyo  propio  análisis  político"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.15442273131374676\n",
+      "model : LinearSVM , score = 0.16483321148477884\n",
+      "model : RBFSVM , score = 0.17021238841374622\n",
+      "model : DecisionTree , score = 0.12652932765832817\n",
+      "model : RandomForest , score = 0.10259509639472521\n",
+      "model : AdaBoost , score = 0.1258932983112584\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 75_64+75_108_68+100_38+100_37_64+100_37_108_68+.csv\n",
+      "Rules(classes) number : 5\n",
+      "Words(features) number : 4\n",
+      "Records number : 40"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.197488</td>\n",
+       "      <td>incluso</td>\n",
+       "      <td>llegar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>poner</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.197488</td>\n",
+       "      <td>incluso</td>\n",
+       "      <td>llegar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>poner</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.210048</td>\n",
+       "      <td>incluso</td>\n",
+       "      <td>llegar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>poner</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>0.197488</td>\n",
+       "      <td>incluso</td>\n",
+       "      <td>llegar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>poner</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>4</td>\n",
+       "      <td>0.197488</td>\n",
+       "      <td>incluso</td>\n",
+       "      <td>llegar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>poner</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight    word1   word2 word3  word4\n",
+       "0     0  0.197488  incluso  llegar     a  poner\n",
+       "1     1  0.197488  incluso  llegar     a  poner\n",
+       "2     2  0.210048  incluso  llegar     a  poner\n",
+       "3     3  0.197488  incluso  llegar     a  poner\n",
+       "4     4  0.197488  incluso  llegar     a  poner"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.04963963207633959\n",
+      "model : LinearSVM , score = 0.2079680095233451\n",
+      "model : RBFSVM , score = 0.2079680095233451\n",
+      "model : DecisionTree , score = 0.052796653155487816\n",
+      "model : RandomForest , score = 0.052796653155487816\n",
+      "model : AdaBoost , score = 0.052796653155487816\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 40_64+40_91+40_108_68+37_67+37_63_64+37_63_91+37_63_108_68+.csv\n",
+      "Rules(classes) number : 7\n",
+      "Words(features) number : 4\n",
+      "Records number : 91"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.144200</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>tratar</td>\n",
+       "      <td>de</td>\n",
+       "      <td>mejorar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.141673</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>tratar</td>\n",
+       "      <td>de</td>\n",
+       "      <td>mejorar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.139854</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>tratar</td>\n",
+       "      <td>de</td>\n",
+       "      <td>mejorar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>0.144200</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>tratar</td>\n",
+       "      <td>de</td>\n",
+       "      <td>mejorar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>4</td>\n",
+       "      <td>0.144200</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>tratar</td>\n",
+       "      <td>de</td>\n",
+       "      <td>mejorar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>5</th>\n",
+       "      <td>5</td>\n",
+       "      <td>0.141673</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>tratar</td>\n",
+       "      <td>de</td>\n",
+       "      <td>mejorar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>6</th>\n",
+       "      <td>6</td>\n",
+       "      <td>0.144200</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>tratar</td>\n",
+       "      <td>de</td>\n",
+       "      <td>mejorar</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight  word1   word2 word3    word4\n",
+       "0     0  0.144200  haber  tratar    de  mejorar\n",
+       "1     1  0.141673  haber  tratar    de  mejorar\n",
+       "2     2  0.139854  haber  tratar    de  mejorar\n",
+       "3     3  0.144200  haber  tratar    de  mejorar\n",
+       "4     4  0.144200  haber  tratar    de  mejorar\n",
+       "5     5  0.141673  haber  tratar    de  mejorar\n",
+       "6     6  0.144200  haber  tratar    de  mejorar"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.04232042772474502\n",
+      "model : LinearSVM , score = 0.13311660024331154\n",
+      "model : RBFSVM , score = 0.13311660024331154\n",
+      "model : DecisionTree , score = 0.0\n",
+      "model : RandomForest , score = 0.02119754231305326\n",
+      "model : AdaBoost , score = 0.0653773982389182\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 76_64+76_108_68+75_67+75_63_64+75_63_108_68+100_40_64+100_40_108_68+100_37_67+100_37_63_64+100_37_63_108_68+.csv\n",
+      "Rules(classes) number : 10\n",
+      "Words(features) number : 5\n",
+      "Records number : 140"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "      <th>word5</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.1</td>\n",
+       "      <td>ahora</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>volver</td>\n",
+       "      <td>a</td>\n",
+       "      <td>ser</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.1</td>\n",
+       "      <td>ahora</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>volver</td>\n",
+       "      <td>a</td>\n",
+       "      <td>ser</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.1</td>\n",
+       "      <td>ahora</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>volver</td>\n",
+       "      <td>a</td>\n",
+       "      <td>ser</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>0.1</td>\n",
+       "      <td>ahora</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>volver</td>\n",
+       "      <td>a</td>\n",
+       "      <td>ser</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>4</td>\n",
+       "      <td>0.1</td>\n",
+       "      <td>ahora</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>volver</td>\n",
+       "      <td>a</td>\n",
+       "      <td>ser</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>5</th>\n",
+       "      <td>5</td>\n",
+       "      <td>0.1</td>\n",
+       "      <td>ahora</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>volver</td>\n",
+       "      <td>a</td>\n",
+       "      <td>ser</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>6</th>\n",
+       "      <td>6</td>\n",
+       "      <td>0.1</td>\n",
+       "      <td>ahora</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>volver</td>\n",
+       "      <td>a</td>\n",
+       "      <td>ser</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7</th>\n",
+       "      <td>7</td>\n",
+       "      <td>0.1</td>\n",
+       "      <td>ahora</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>volver</td>\n",
+       "      <td>a</td>\n",
+       "      <td>ser</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>8</th>\n",
+       "      <td>8</td>\n",
+       "      <td>0.1</td>\n",
+       "      <td>ahora</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>volver</td>\n",
+       "      <td>a</td>\n",
+       "      <td>ser</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>9</th>\n",
+       "      <td>9</td>\n",
+       "      <td>0.1</td>\n",
+       "      <td>ahora</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>volver</td>\n",
+       "      <td>a</td>\n",
+       "      <td>ser</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule  weight  word1  word2   word3 word4 word5\n",
+       "0     0     0.1  ahora  haber  volver     a   ser\n",
+       "1     1     0.1  ahora  haber  volver     a   ser\n",
+       "2     2     0.1  ahora  haber  volver     a   ser\n",
+       "3     3     0.1  ahora  haber  volver     a   ser\n",
+       "4     4     0.1  ahora  haber  volver     a   ser\n",
+       "5     5     0.1  ahora  haber  volver     a   ser\n",
+       "6     6     0.1  ahora  haber  volver     a   ser\n",
+       "7     7     0.1  ahora  haber  volver     a   ser\n",
+       "8     8     0.1  ahora  haber  volver     a   ser\n",
+       "9     9     0.1  ahora  haber  volver     a   ser"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.029109801540275218\n",
+      "model : LinearSVM , score = 0.10136400242342919\n",
+      "model : RBFSVM , score = 0.10136400242342919\n",
+      "model : DecisionTree , score = 0.0\n",
+      "model : RandomForest , score = 0.0\n",
+      "model : AdaBoost , score = 0.05738313227876574\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 99+113_108_1+.csv\n",
+      "Rules(classes) number : 2\n",
+      "Words(features) number : 3\n",
+      "Records number : 110"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.523369</td>\n",
+       "      <td>18</td>\n",
+       "      <td>de</td>\n",
+       "      <td>noviembre</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.476631</td>\n",
+       "      <td>18</td>\n",
+       "      <td>de</td>\n",
+       "      <td>noviembre</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight  word1 word2      word3\n",
+       "0     0  0.523369     18    de  noviembre\n",
+       "1     1  0.476631     18    de  noviembre"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.46647297102853236\n",
+      "model : LinearSVM , score = 0.4616156736822202\n",
+      "model : RBFSVM , score = 0.4927045780339101\n",
+      "model : DecisionTree , score = 0.3703357021742049\n",
+      "model : RandomForest , score = 0.29095544220379604\n",
+      "model : AdaBoost , score = 0.3344475953181826\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 15+1_63_29+.csv\n",
+      "Rules(classes) number : 2\n",
+      "Words(features) number : 3\n",
+      "Records number : 12"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.522456</td>\n",
+       "      <td>norma</td>\n",
+       "      <td>armonizar</td>\n",
+       "      <td>común</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.477544</td>\n",
+       "      <td>norma</td>\n",
+       "      <td>armonizar</td>\n",
+       "      <td>común</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight  word1      word2  word3\n",
+       "0     0  0.522456  norma  armonizar  común\n",
+       "1     1  0.477544  norma  armonizar  común"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.17453341368671002\n",
+      "model : LinearSVM , score = 0.5096781499837646\n",
+      "model : RBFSVM , score = 0.5096781499837646\n",
+      "model : DecisionTree , score = 0.17453341368671002\n",
+      "model : RandomForest , score = 0.17453341368671002\n",
+      "model : AdaBoost , score = 0.17453341368671002\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 79+84_63+100_42+100_45_63+100_47_63+.csv\n",
+      "Rules(classes) number : 5\n",
+      "Words(features) number : 4\n",
+      "Records number : 15"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.227411</td>\n",
+       "      <td>no</td>\n",
+       "      <td>lo</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>ser</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.194775</td>\n",
+       "      <td>no</td>\n",
+       "      <td>lo</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>ser</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.210592</td>\n",
+       "      <td>no</td>\n",
+       "      <td>lo</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>ser</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>0.178748</td>\n",
+       "      <td>no</td>\n",
+       "      <td>lo</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>ser</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>4</td>\n",
+       "      <td>0.188474</td>\n",
+       "      <td>no</td>\n",
+       "      <td>lo</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>ser</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight word1 word2  word3 word4\n",
+       "0     0  0.227411    no    lo  haber   ser\n",
+       "1     1  0.194775    no    lo  haber   ser\n",
+       "2     2  0.210592    no    lo  haber   ser\n",
+       "3     3  0.178748    no    lo  haber   ser\n",
+       "4     4  0.188474    no    lo  haber   ser"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.0\n",
+      "model : LinearSVM , score = 0.12847496543948675\n",
+      "model : RBFSVM , score = 0.12847496543948675\n",
+      "model : DecisionTree , score = 0.0\n",
+      "model : RandomForest , score = 0.0\n",
+      "model : AdaBoost , score = 0.0\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 22+12_30+12_100_63+113_1_30+113_1_100_63+.csv\n",
+      "Rules(classes) number : 5\n",
+      "Words(features) number : 4\n",
+      "Records number : 30"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.203416</td>\n",
+       "      <td>tres</td>\n",
+       "      <td>tema</td>\n",
+       "      <td>estrechamente</td>\n",
+       "      <td>interrelacionar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.199146</td>\n",
+       "      <td>tres</td>\n",
+       "      <td>tema</td>\n",
+       "      <td>estrechamente</td>\n",
+       "      <td>interrelacionar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.199146</td>\n",
+       "      <td>tres</td>\n",
+       "      <td>tema</td>\n",
+       "      <td>estrechamente</td>\n",
+       "      <td>interrelacionar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>0.199146</td>\n",
+       "      <td>tres</td>\n",
+       "      <td>tema</td>\n",
+       "      <td>estrechamente</td>\n",
+       "      <td>interrelacionar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>4</td>\n",
+       "      <td>0.199146</td>\n",
+       "      <td>tres</td>\n",
+       "      <td>tema</td>\n",
+       "      <td>estrechamente</td>\n",
+       "      <td>interrelacionar</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight word1 word2          word3            word4\n",
+       "0     0  0.203416  tres  tema  estrechamente  interrelacionar\n",
+       "1     1  0.199146  tres  tema  estrechamente  interrelacionar\n",
+       "2     2  0.199146  tres  tema  estrechamente  interrelacionar\n",
+       "3     3  0.199146  tres  tema  estrechamente  interrelacionar\n",
+       "4     4  0.199146  tres  tema  estrechamente  interrelacionar"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.1325296415453355\n",
+      "model : LinearSVM , score = 0.19959612235054666\n",
+      "model : RBFSVM , score = 0.19959612235054666\n",
+      "model : DecisionTree , score = 0.06596999538879848\n",
+      "model : RandomForest , score = 0.06596999538879848\n",
+      "model : AdaBoost , score = 0.06596999538879848\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 21_64+21_108_68+10_30_64+10_30_108_68+10_100_67+10_100_63_64+10_100_63_108_68+11_30_64+11_30_108_68+11_100_67+11_100_63_64+11_100_63_108_68+33_1_30_64+33_1_30_108_68+33_1_100_67+33_1_100_63_64+33_1_100_63_108_68+.csv\n",
+      "Rules(classes) number : 17\n",
+      "Words(features) number : 6\n",
+      "Records number : 34"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "      <th>word5</th>\n",
+       "      <th>word6</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.061179</td>\n",
+       "      <td>nuestro</td>\n",
+       "      <td>diputado</td>\n",
+       "      <td>más</td>\n",
+       "      <td>acostumbrar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>intervenir</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.057171</td>\n",
+       "      <td>nuestro</td>\n",
+       "      <td>diputado</td>\n",
+       "      <td>más</td>\n",
+       "      <td>acostumbrar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>intervenir</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.057728</td>\n",
+       "      <td>nuestro</td>\n",
+       "      <td>diputado</td>\n",
+       "      <td>más</td>\n",
+       "      <td>acostumbrar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>intervenir</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>0.057728</td>\n",
+       "      <td>nuestro</td>\n",
+       "      <td>diputado</td>\n",
+       "      <td>más</td>\n",
+       "      <td>acostumbrar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>intervenir</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>4</td>\n",
+       "      <td>0.057728</td>\n",
+       "      <td>nuestro</td>\n",
+       "      <td>diputado</td>\n",
+       "      <td>más</td>\n",
+       "      <td>acostumbrar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>intervenir</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>5</th>\n",
+       "      <td>5</td>\n",
+       "      <td>0.057728</td>\n",
+       "      <td>nuestro</td>\n",
+       "      <td>diputado</td>\n",
+       "      <td>más</td>\n",
+       "      <td>acostumbrar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>intervenir</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>6</th>\n",
+       "      <td>6</td>\n",
+       "      <td>0.057728</td>\n",
+       "      <td>nuestro</td>\n",
+       "      <td>diputado</td>\n",
+       "      <td>más</td>\n",
+       "      <td>acostumbrar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>intervenir</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7</th>\n",
+       "      <td>7</td>\n",
+       "      <td>0.060875</td>\n",
+       "      <td>nuestro</td>\n",
+       "      <td>diputado</td>\n",
+       "      <td>más</td>\n",
+       "      <td>acostumbrar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>intervenir</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>8</th>\n",
+       "      <td>8</td>\n",
+       "      <td>0.060875</td>\n",
+       "      <td>nuestro</td>\n",
+       "      <td>diputado</td>\n",
+       "      <td>más</td>\n",
+       "      <td>acostumbrar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>intervenir</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>9</th>\n",
+       "      <td>9</td>\n",
+       "      <td>0.060875</td>\n",
+       "      <td>nuestro</td>\n",
+       "      <td>diputado</td>\n",
+       "      <td>más</td>\n",
+       "      <td>acostumbrar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>intervenir</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>10</th>\n",
+       "      <td>10</td>\n",
+       "      <td>0.060875</td>\n",
+       "      <td>nuestro</td>\n",
+       "      <td>diputado</td>\n",
+       "      <td>más</td>\n",
+       "      <td>acostumbrar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>intervenir</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>11</th>\n",
+       "      <td>11</td>\n",
+       "      <td>0.060875</td>\n",
+       "      <td>nuestro</td>\n",
+       "      <td>diputado</td>\n",
+       "      <td>más</td>\n",
+       "      <td>acostumbrar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>intervenir</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>12</th>\n",
+       "      <td>12</td>\n",
+       "      <td>0.057728</td>\n",
+       "      <td>nuestro</td>\n",
+       "      <td>diputado</td>\n",
+       "      <td>más</td>\n",
+       "      <td>acostumbrar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>intervenir</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>13</th>\n",
+       "      <td>13</td>\n",
+       "      <td>0.057728</td>\n",
+       "      <td>nuestro</td>\n",
+       "      <td>diputado</td>\n",
+       "      <td>más</td>\n",
+       "      <td>acostumbrar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>intervenir</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>14</th>\n",
+       "      <td>14</td>\n",
+       "      <td>0.057728</td>\n",
+       "      <td>nuestro</td>\n",
+       "      <td>diputado</td>\n",
+       "      <td>más</td>\n",
+       "      <td>acostumbrar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>intervenir</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>15</th>\n",
+       "      <td>15</td>\n",
+       "      <td>0.057728</td>\n",
+       "      <td>nuestro</td>\n",
+       "      <td>diputado</td>\n",
+       "      <td>más</td>\n",
+       "      <td>acostumbrar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>intervenir</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>16</th>\n",
+       "      <td>16</td>\n",
+       "      <td>0.057728</td>\n",
+       "      <td>nuestro</td>\n",
+       "      <td>diputado</td>\n",
+       "      <td>más</td>\n",
+       "      <td>acostumbrar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>intervenir</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "    rule    weight    word1     word2 word3        word4 word5       word6\n",
+       "0      0  0.061179  nuestro  diputado   más  acostumbrar     a  intervenir\n",
+       "1      1  0.057171  nuestro  diputado   más  acostumbrar     a  intervenir\n",
+       "2      2  0.057728  nuestro  diputado   más  acostumbrar     a  intervenir\n",
+       "3      3  0.057728  nuestro  diputado   más  acostumbrar     a  intervenir\n",
+       "4      4  0.057728  nuestro  diputado   más  acostumbrar     a  intervenir\n",
+       "5      5  0.057728  nuestro  diputado   más  acostumbrar     a  intervenir\n",
+       "6      6  0.057728  nuestro  diputado   más  acostumbrar     a  intervenir\n",
+       "7      7  0.060875  nuestro  diputado   más  acostumbrar     a  intervenir\n",
+       "8      8  0.060875  nuestro  diputado   más  acostumbrar     a  intervenir\n",
+       "9      9  0.060875  nuestro  diputado   más  acostumbrar     a  intervenir\n",
+       "10    10  0.060875  nuestro  diputado   más  acostumbrar     a  intervenir\n",
+       "11    11  0.060875  nuestro  diputado   más  acostumbrar     a  intervenir\n",
+       "12    12  0.057728  nuestro  diputado   más  acostumbrar     a  intervenir\n",
+       "13    13  0.057728  nuestro  diputado   más  acostumbrar     a  intervenir\n",
+       "14    14  0.057728  nuestro  diputado   más  acostumbrar     a  intervenir\n",
+       "15    15  0.057728  nuestro  diputado   más  acostumbrar     a  intervenir\n",
+       "16    16  0.057728  nuestro  diputado   más  acostumbrar     a  intervenir"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.061179375528249785\n",
+      "model : LinearSVM , score = 0.061179375528249785\n",
+      "model : RBFSVM , score = 0.061179375528249785\n",
+      "model : DecisionTree , score = 0.061179375528249785\n",
+      "model : RandomForest , score = 0.05772767690892923\n",
+      "model : AdaBoost , score = 0.061179375528249785\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 30_108_52+100_63_108_52+.csv\n",
+      "Rules(classes) number : 2\n",
+      "Words(features) number : 5\n",
+      "Records number : 4"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "/home/aboelhamd/anaconda3/lib/python3.7/site-packages/sklearn/naive_bayes.py:436: RuntimeWarning: divide by zero encountered in log\n",
+      "  n_ij = - 0.5 * np.sum(np.log(2. * np.pi * self.sigma_[i, :]))\n",
+      "/home/aboelhamd/anaconda3/lib/python3.7/site-packages/sklearn/naive_bayes.py:438: RuntimeWarning: invalid value encountered in true_divide\n",
+      "  (self.sigma_[i, :]), 1)\n"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "      <th>word5</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.5</td>\n",
+       "      <td>demasiado</td>\n",
+       "      <td>asustar</td>\n",
+       "      <td>para</td>\n",
+       "      <td>hacer</td>\n",
+       "      <td>lo</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.5</td>\n",
+       "      <td>demasiado</td>\n",
+       "      <td>asustar</td>\n",
+       "      <td>para</td>\n",
+       "      <td>hacer</td>\n",
+       "      <td>lo</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule  weight      word1    word2 word3  word4 word5\n",
+       "0     0     0.5  demasiado  asustar  para  hacer    lo\n",
+       "1     1     0.5  demasiado  asustar  para  hacer    lo"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.0\n",
+      "model : LinearSVM , score = 0.0\n",
+      "model : RBFSVM , score = 0.0\n",
+      "model : DecisionTree , score = 0.0\n",
+      "model : RandomForest , score = 0.0\n",
+      "model : AdaBoost , score = 0.0\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 85+100_52+.csv\n",
+      "Rules(classes) number : 2\n",
+      "Words(features) number : 3\n",
+      "Records number : 18"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.499061</td>\n",
+       "      <td>no</td>\n",
+       "      <td>utilizar</td>\n",
+       "      <td>lo</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.500939</td>\n",
+       "      <td>no</td>\n",
+       "      <td>utilizar</td>\n",
+       "      <td>lo</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight word1     word2 word3\n",
+       "0     0  0.499061    no  utilizar    lo\n",
+       "1     1  0.500939    no  utilizar    lo"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.2236238975351392\n",
+      "model : LinearSVM , score = 0.44156793810892925\n",
+      "model : RBFSVM , score = 0.44156793810892925\n",
+      "model : DecisionTree , score = 0.2236238975351392\n",
+      "model : RandomForest , score = 0.2236238975351392\n",
+      "model : AdaBoost , score = 0.2236238975351392\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 66+108_33_68+.csv\n",
+      "Rules(classes) number : 2\n",
+      "Words(features) number : 3\n",
+      "Records number : 308"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.492659</td>\n",
+       "      <td>a</td>\n",
+       "      <td>el</td>\n",
+       "      <td>señalar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.507341</td>\n",
+       "      <td>a</td>\n",
+       "      <td>el</td>\n",
+       "      <td>señalar</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight word1 word2    word3\n",
+       "0     0  0.492659     a    el  señalar\n",
+       "1     1  0.507341     a    el  señalar"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.461297191301524\n",
+      "model : LinearSVM , score = 0.46970196445060763\n",
+      "model : RBFSVM , score = 0.28830592251804876\n",
+      "model : DecisionTree , score = 0.4122573664704467\n",
+      "model : RandomForest , score = 0.36884414076967253\n",
+      "model : AdaBoost , score = 0.37300673572156284\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 42+47_63+.csv\n",
+      "Rules(classes) number : 2\n",
+      "Words(features) number : 3\n",
+      "Records number : 244"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.499704</td>\n",
+       "      <td>lo</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>ser</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.500296</td>\n",
+       "      <td>lo</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>ser</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight word1  word2 word3\n",
+       "0     0  0.499704    lo  haber   ser\n",
+       "1     1  0.500296    lo  haber   ser"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.38677502238711753\n",
+      "model : LinearSVM , score = 0.4636285737988887\n",
+      "model : RBFSVM , score = 0.3184207452419576\n",
+      "model : DecisionTree , score = 0.3351521422881191\n",
+      "model : RandomForest , score = 0.35993603057560336\n",
+      "model : AdaBoost , score = 0.3624322759665186\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 21+10_30+10_100_63+11_30+11_100_63+33_1_30+33_1_100_63+.csv\n",
+      "Rules(classes) number : 7\n",
+      "Words(features) number : 4\n",
+      "Records number : 1120"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.144982</td>\n",
+       "      <td>uno</td>\n",
+       "      <td>sistema</td>\n",
+       "      <td>totalmente</td>\n",
+       "      <td>descentralizar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.142143</td>\n",
+       "      <td>uno</td>\n",
+       "      <td>sistema</td>\n",
+       "      <td>totalmente</td>\n",
+       "      <td>descentralizar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.142143</td>\n",
+       "      <td>uno</td>\n",
+       "      <td>sistema</td>\n",
+       "      <td>totalmente</td>\n",
+       "      <td>descentralizar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>0.143223</td>\n",
+       "      <td>uno</td>\n",
+       "      <td>sistema</td>\n",
+       "      <td>totalmente</td>\n",
+       "      <td>descentralizar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>4</td>\n",
+       "      <td>0.143223</td>\n",
+       "      <td>uno</td>\n",
+       "      <td>sistema</td>\n",
+       "      <td>totalmente</td>\n",
+       "      <td>descentralizar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>5</th>\n",
+       "      <td>5</td>\n",
+       "      <td>0.142143</td>\n",
+       "      <td>uno</td>\n",
+       "      <td>sistema</td>\n",
+       "      <td>totalmente</td>\n",
+       "      <td>descentralizar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>6</th>\n",
+       "      <td>6</td>\n",
+       "      <td>0.142143</td>\n",
+       "      <td>uno</td>\n",
+       "      <td>sistema</td>\n",
+       "      <td>totalmente</td>\n",
+       "      <td>descentralizar</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight word1    word2       word3           word4\n",
+       "0     0  0.144982   uno  sistema  totalmente  descentralizar\n",
+       "1     1  0.142143   uno  sistema  totalmente  descentralizar\n",
+       "2     2  0.142143   uno  sistema  totalmente  descentralizar\n",
+       "3     3  0.143223   uno  sistema  totalmente  descentralizar\n",
+       "4     4  0.143223   uno  sistema  totalmente  descentralizar\n",
+       "5     5  0.142143   uno  sistema  totalmente  descentralizar\n",
+       "6     6  0.142143   uno  sistema  totalmente  descentralizar"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.1177903737614213\n",
+      "model : LinearSVM , score = 0.1120123893124631\n",
+      "model : RBFSVM , score = 0.1436045515362441\n",
+      "model : DecisionTree , score = 0.05609071186648488\n",
+      "model : RandomForest , score = 0.04315566291476371\n",
+      "model : AdaBoost , score = 0.08045588756829092\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 96_1+99_108_12+99_108_113_1+113_108_1_108_12+113_108_1_108_113_1+.csv\n",
+      "Rules(classes) number : 5\n",
+      "Words(features) number : 6\n",
+      "Records number : 20"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "      <th>word5</th>\n",
+       "      <th>word6</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.241524</td>\n",
+       "      <td>15</td>\n",
+       "      <td>de</td>\n",
+       "      <td>diciembre</td>\n",
+       "      <td>de</td>\n",
+       "      <td>2006</td>\n",
+       "      <td>continuación</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.201304</td>\n",
+       "      <td>15</td>\n",
+       "      <td>de</td>\n",
+       "      <td>diciembre</td>\n",
+       "      <td>de</td>\n",
+       "      <td>2006</td>\n",
+       "      <td>continuación</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.201304</td>\n",
+       "      <td>15</td>\n",
+       "      <td>de</td>\n",
+       "      <td>diciembre</td>\n",
+       "      <td>de</td>\n",
+       "      <td>2006</td>\n",
+       "      <td>continuación</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>0.177934</td>\n",
+       "      <td>15</td>\n",
+       "      <td>de</td>\n",
+       "      <td>diciembre</td>\n",
+       "      <td>de</td>\n",
+       "      <td>2006</td>\n",
+       "      <td>continuación</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>4</td>\n",
+       "      <td>0.177934</td>\n",
+       "      <td>15</td>\n",
+       "      <td>de</td>\n",
+       "      <td>diciembre</td>\n",
+       "      <td>de</td>\n",
+       "      <td>2006</td>\n",
+       "      <td>continuación</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight  word1 word2      word3 word4  word5         word6\n",
+       "0     0  0.241524     15    de  diciembre    de   2006  continuación\n",
+       "1     1  0.201304     15    de  diciembre    de   2006  continuación\n",
+       "2     2  0.201304     15    de  diciembre    de   2006  continuación\n",
+       "3     3  0.177934     15    de  diciembre    de   2006  continuación\n",
+       "4     4  0.177934     15    de  diciembre    de   2006  continuación"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.0\n",
+      "model : LinearSVM , score = 0.2413839808040335\n",
+      "model : RBFSVM , score = 0.2413839808040335\n",
+      "model : DecisionTree , score = 0.0\n",
+      "model : RandomForest , score = 0.0\n",
+      "model : AdaBoost , score = 0.0\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 47_64+47_108_68+.csv\n",
+      "Rules(classes) number : 2\n",
+      "Words(features) number : 4\n",
+      "Records number : 4"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.5</td>\n",
+       "      <td>lo</td>\n",
+       "      <td>llegar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>saber</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.5</td>\n",
+       "      <td>lo</td>\n",
+       "      <td>llegar</td>\n",
+       "      <td>a</td>\n",
+       "      <td>saber</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule  weight word1   word2 word3  word4\n",
+       "0     0     0.5    lo  llegar     a  saber\n",
+       "1     1     0.5    lo  llegar     a  saber"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.5\n",
+      "model : LinearSVM , score = 0.5\n",
+      "model : RBFSVM , score = 0.5\n",
+      "model : DecisionTree , score = 0.5\n",
+      "model : RandomForest , score = 0.5\n",
+      "model : AdaBoost , score = 0.5\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 21+10_30+10_110_63+11_30+11_110_63+33_1_30+33_1_110_63+.csv\n",
+      "Rules(classes) number : 7\n",
+      "Words(features) number : 4\n",
+      "Records number : 287"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "/home/aboelhamd/anaconda3/lib/python3.7/site-packages/sklearn/naive_bayes.py:436: RuntimeWarning: divide by zero encountered in log\n",
+      "  n_ij = - 0.5 * np.sum(np.log(2. * np.pi * self.sigma_[i, :]))\n",
+      "/home/aboelhamd/anaconda3/lib/python3.7/site-packages/sklearn/naive_bayes.py:438: RuntimeWarning: invalid value encountered in true_divide\n",
+      "  (self.sigma_[i, :]), 1)\n"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.132379</td>\n",
+       "      <td>suyo</td>\n",
+       "      <td>actuación</td>\n",
+       "      <td>tan</td>\n",
+       "      <td>decidir</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.142428</td>\n",
+       "      <td>suyo</td>\n",
+       "      <td>actuación</td>\n",
+       "      <td>tan</td>\n",
+       "      <td>decidir</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.142428</td>\n",
+       "      <td>suyo</td>\n",
+       "      <td>actuación</td>\n",
+       "      <td>tan</td>\n",
+       "      <td>decidir</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>0.148955</td>\n",
+       "      <td>suyo</td>\n",
+       "      <td>actuación</td>\n",
+       "      <td>tan</td>\n",
+       "      <td>decidir</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>4</td>\n",
+       "      <td>0.148955</td>\n",
+       "      <td>suyo</td>\n",
+       "      <td>actuación</td>\n",
+       "      <td>tan</td>\n",
+       "      <td>decidir</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>5</th>\n",
+       "      <td>5</td>\n",
+       "      <td>0.142428</td>\n",
+       "      <td>suyo</td>\n",
+       "      <td>actuación</td>\n",
+       "      <td>tan</td>\n",
+       "      <td>decidir</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>6</th>\n",
+       "      <td>6</td>\n",
+       "      <td>0.142428</td>\n",
+       "      <td>suyo</td>\n",
+       "      <td>actuación</td>\n",
+       "      <td>tan</td>\n",
+       "      <td>decidir</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight word1      word2 word3    word4\n",
+       "0     0  0.132379  suyo  actuación   tan  decidir\n",
+       "1     1  0.142428  suyo  actuación   tan  decidir\n",
+       "2     2  0.142428  suyo  actuación   tan  decidir\n",
+       "3     3  0.148955  suyo  actuación   tan  decidir\n",
+       "4     4  0.148955  suyo  actuación   tan  decidir\n",
+       "5     5  0.142428  suyo  actuación   tan  decidir\n",
+       "6     6  0.142428  suyo  actuación   tan  decidir"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.12117854955889858\n",
+      "model : LinearSVM , score = 0.11058929905981565\n",
+      "model : RBFSVM , score = 0.13765637162946023\n",
+      "model : DecisionTree , score = 0.05708629776372969\n",
+      "model : RandomForest , score = 0.013768958717064288\n",
+      "model : AdaBoost , score = 0.048194050630391144\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 70+37_69+.csv\n",
+      "Rules(classes) number : 2\n",
+      "Words(features) number : 2\n",
+      "Records number : 1756"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.514636</td>\n",
+       "      <td>estar</td>\n",
+       "      <td>examinar</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.485364</td>\n",
+       "      <td>estar</td>\n",
+       "      <td>examinar</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight  word1     word2\n",
+       "0     0  0.514636  estar  examinar\n",
+       "1     1  0.485364  estar  examinar"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.5092376080873654\n",
+      "model : LinearSVM , score = 0.5155477821955644\n",
+      "model : RBFSVM , score = 0.4142580693139405\n",
+      "model : DecisionTree , score = 0.47187299992698806\n",
+      "model : RandomForest , score = 0.4381772239282214\n",
+      "model : AdaBoost , score = 0.4576496467893957\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 12+113_1+.csv\n",
+      "Rules(classes) number : 2\n",
+      "Words(features) number : 2\n",
+      "Records number : 6242"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.5</td>\n",
+       "      <td>seis</td>\n",
+       "      <td>mes</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.5</td>\n",
+       "      <td>seis</td>\n",
+       "      <td>mes</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule  weight word1 word2\n",
+       "0     0     0.5  seis   mes\n",
+       "1     1     0.5  seis   mes"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.5023250356357817\n",
+      "model : LinearSVM , score = 0.49487183011875113\n",
+      "model : RBFSVM , score = 0.3587189323620225\n",
+      "model : DecisionTree , score = 0.4876632082487883\n",
+      "model : RandomForest , score = 0.4593944237711818\n",
+      "model : AdaBoost , score = 0.4652437067509656\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 22+12_30+12_100_29+113_1_30+113_1_100_29+.csv\n",
+      "Rules(classes) number : 5\n",
+      "Words(features) number : 4\n",
+      "Records number : 60"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.208552</td>\n",
+       "      <td>50</td>\n",
+       "      <td>país</td>\n",
+       "      <td>más</td>\n",
+       "      <td>pobre</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.201767</td>\n",
+       "      <td>50</td>\n",
+       "      <td>país</td>\n",
+       "      <td>más</td>\n",
+       "      <td>pobre</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.193957</td>\n",
+       "      <td>50</td>\n",
+       "      <td>país</td>\n",
+       "      <td>más</td>\n",
+       "      <td>pobre</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>0.201767</td>\n",
+       "      <td>50</td>\n",
+       "      <td>país</td>\n",
+       "      <td>más</td>\n",
+       "      <td>pobre</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>4</td>\n",
+       "      <td>0.193957</td>\n",
+       "      <td>50</td>\n",
+       "      <td>país</td>\n",
+       "      <td>más</td>\n",
+       "      <td>pobre</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight word1 word2 word3  word4\n",
+       "0     0  0.208552    50  país   más  pobre\n",
+       "1     1  0.201767    50  país   más  pobre\n",
+       "2     2  0.193957    50  país   más  pobre\n",
+       "3     3  0.201767    50  país   más  pobre\n",
+       "4     4  0.193957    50  país   más  pobre"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.03243725296195384\n",
+      "model : LinearSVM , score = 0.20664318571254572\n",
+      "model : RBFSVM , score = 0.20664318571254572\n",
+      "model : DecisionTree , score = 0.0\n",
+      "model : RandomForest , score = 0.0\n",
+      "model : AdaBoost , score = 0.03312666534301269\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 15+14_29+1_29_29+.csv\n",
+      "Rules(classes) number : 3\n",
+      "Words(features) number : 3\n",
+      "Records number : 777"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.336066</td>\n",
+       "      <td>concepto</td>\n",
+       "      <td>jurídico</td>\n",
+       "      <td>indeterminado</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.331967</td>\n",
+       "      <td>concepto</td>\n",
+       "      <td>jurídico</td>\n",
+       "      <td>indeterminado</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.331967</td>\n",
+       "      <td>concepto</td>\n",
+       "      <td>jurídico</td>\n",
+       "      <td>indeterminado</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight     word1     word2          word3\n",
+       "0     0  0.336066  concepto  jurídico  indeterminado\n",
+       "1     1  0.331967  concepto  jurídico  indeterminado\n",
+       "2     2  0.331967  concepto  jurídico  indeterminado"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.2936050137852968\n",
+      "model : LinearSVM , score = 0.29726552361270925\n",
+      "model : RBFSVM , score = 0.33627700140717914\n",
+      "model : DecisionTree , score = 0.2752638365628531\n",
+      "model : RandomForest , score = 0.16709213403780793\n",
+      "model : AdaBoost , score = 0.1985863322341775\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 84+100_60+100_47+.csv\n",
+      "Rules(classes) number : 3\n",
+      "Words(features) number : 3\n",
+      "Records number : 12"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.328663</td>\n",
+       "      <td>no</td>\n",
+       "      <td>lo</td>\n",
+       "      <td>poder</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.345980</td>\n",
+       "      <td>no</td>\n",
+       "      <td>lo</td>\n",
+       "      <td>poder</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.325357</td>\n",
+       "      <td>no</td>\n",
+       "      <td>lo</td>\n",
+       "      <td>poder</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight word1 word2  word3\n",
+       "0     0  0.328663    no    lo  poder\n",
+       "1     1  0.345980    no    lo  poder\n",
+       "2     2  0.325357    no    lo  poder"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.16665342755583534\n",
+      "model : LinearSVM , score = 0.3344504145340458\n",
+      "model : RBFSVM , score = 0.3344504145340458\n",
+      "model : DecisionTree , score = 0.16665342755583534\n",
+      "model : RandomForest , score = 0.1655003759407887\n",
+      "model : AdaBoost , score = 0.16665342755583534\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 94+97_108_113+108_95+108_98_108_113+108_33_96+108_33_99_108_113+108_33_113_108_1_108_113+.csv\n",
+      "Rules(classes) number : 7\n",
+      "Words(features) number : 7\n",
+      "Records number : 273"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "      <th>word5</th>\n",
+       "      <th>word6</th>\n",
+       "      <th>word7</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.162927</td>\n",
+       "      <td>a</td>\n",
+       "      <td>el</td>\n",
+       "      <td>27</td>\n",
+       "      <td>de</td>\n",
+       "      <td>abril</td>\n",
+       "      <td>de</td>\n",
+       "      <td>2000</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.136174</td>\n",
+       "      <td>a</td>\n",
+       "      <td>el</td>\n",
+       "      <td>27</td>\n",
+       "      <td>de</td>\n",
+       "      <td>abril</td>\n",
+       "      <td>de</td>\n",
+       "      <td>2000</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.154397</td>\n",
+       "      <td>a</td>\n",
+       "      <td>el</td>\n",
+       "      <td>27</td>\n",
+       "      <td>de</td>\n",
+       "      <td>abril</td>\n",
+       "      <td>de</td>\n",
+       "      <td>2000</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>0.130163</td>\n",
+       "      <td>a</td>\n",
+       "      <td>el</td>\n",
+       "      <td>27</td>\n",
+       "      <td>de</td>\n",
+       "      <td>abril</td>\n",
+       "      <td>de</td>\n",
+       "      <td>2000</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>4</td>\n",
+       "      <td>0.160826</td>\n",
+       "      <td>a</td>\n",
+       "      <td>el</td>\n",
+       "      <td>27</td>\n",
+       "      <td>de</td>\n",
+       "      <td>abril</td>\n",
+       "      <td>de</td>\n",
+       "      <td>2000</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>5</th>\n",
+       "      <td>5</td>\n",
+       "      <td>0.134703</td>\n",
+       "      <td>a</td>\n",
+       "      <td>el</td>\n",
+       "      <td>27</td>\n",
+       "      <td>de</td>\n",
+       "      <td>abril</td>\n",
+       "      <td>de</td>\n",
+       "      <td>2000</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>6</th>\n",
+       "      <td>6</td>\n",
+       "      <td>0.120809</td>\n",
+       "      <td>a</td>\n",
+       "      <td>el</td>\n",
+       "      <td>27</td>\n",
+       "      <td>de</td>\n",
+       "      <td>abril</td>\n",
+       "      <td>de</td>\n",
+       "      <td>2000</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight word1 word2  word3 word4  word5 word6  word7\n",
+       "0     0  0.162927     a    el     27    de  abril    de   2000\n",
+       "1     1  0.136174     a    el     27    de  abril    de   2000\n",
+       "2     2  0.154397     a    el     27    de  abril    de   2000\n",
+       "3     3  0.130163     a    el     27    de  abril    de   2000\n",
+       "4     4  0.160826     a    el     27    de  abril    de   2000\n",
+       "5     5  0.134703     a    el     27    de  abril    de   2000\n",
+       "6     6  0.120809     a    el     27    de  abril    de   2000"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.07477421659088763\n",
+      "model : LinearSVM , score = 0.10496404562553276\n",
+      "model : RBFSVM , score = 0.15827958403992717\n",
+      "model : DecisionTree , score = 0.00769779337779194\n",
+      "model : RandomForest , score = 0.02123447408436888\n",
+      "model : AdaBoost , score = 0.06372186683106934\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 24+14_106_29+1_29_106_29+.csv\n",
+      "Rules(classes) number : 3\n",
+      "Words(features) number : 4\n",
+      "Records number : 789"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.335496</td>\n",
+       "      <td>norma</td>\n",
+       "      <td>claro</td>\n",
+       "      <td>y</td>\n",
+       "      <td>completo</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.331636</td>\n",
+       "      <td>norma</td>\n",
+       "      <td>claro</td>\n",
+       "      <td>y</td>\n",
+       "      <td>completo</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.332868</td>\n",
+       "      <td>norma</td>\n",
+       "      <td>claro</td>\n",
+       "      <td>y</td>\n",
+       "      <td>completo</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight  word1  word2 word3     word4\n",
+       "0     0  0.335496  norma  claro     y  completo\n",
+       "1     1  0.331636  norma  claro     y  completo\n",
+       "2     2  0.332868  norma  claro     y  completo"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.32481851581061016\n",
+      "model : LinearSVM , score = 0.3165909792196462\n",
+      "model : RBFSVM , score = 0.3488859231380752\n",
+      "model : DecisionTree , score = 0.2606002301328052\n",
+      "model : RandomForest , score = 0.20003356990093957\n",
+      "model : AdaBoost , score = 0.22003922441900595\n",
+      "----------------------------------------------\n",
+      "\n",
+      "file name : 60_63+47_40+47_68_63+.csv\n",
+      "Rules(classes) number : 3\n",
+      "Words(features) number : 4\n",
+      "Records number : 6"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>rule</th>\n",
+       "      <th>weight</th>\n",
+       "      <th>word1</th>\n",
+       "      <th>word2</th>\n",
+       "      <th>word3</th>\n",
+       "      <th>word4</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>0.314649</td>\n",
+       "      <td>lo</td>\n",
+       "      <td>poder</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>ser</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>0.342676</td>\n",
+       "      <td>lo</td>\n",
+       "      <td>poder</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>ser</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>0.342676</td>\n",
+       "      <td>lo</td>\n",
+       "      <td>poder</td>\n",
+       "      <td>haber</td>\n",
+       "      <td>ser</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   rule    weight word1  word2  word3 word4\n",
+       "0     0  0.314649    lo  poder  haber   ser\n",
+       "1     1  0.342676    lo  poder  haber   ser\n",
+       "2     2  0.342676    lo  poder  haber   ser"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "model : NaiveBayes , score = 0.3146486853513146\n",
+      "model : LinearSVM , score = 0.34267565732434274\n",
+      "model : RBFSVM , score = 0.34267565732434274\n",
+      "model : DecisionTree , score = 0.34267565732434274\n",
+      "model : RandomForest , score = 0.3146486853513146\n",
+      "model : AdaBoost , score = 0.34267565732434274\n",
+      "----------------------------------------------\n",
+      "\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "/home/aboelhamd/anaconda3/lib/python3.7/site-packages/sklearn/naive_bayes.py:436: RuntimeWarning: divide by zero encountered in log\n",
+      "  n_ij = - 0.5 * np.sum(np.log(2. * np.pi * self.sigma_[i, :]))\n",
+      "/home/aboelhamd/anaconda3/lib/python3.7/site-packages/sklearn/naive_bayes.py:438: RuntimeWarning: invalid value encountered in true_divide\n",
+      "  (self.sigma_[i, :]), 1)\n"
+     ]
+    }
+   ],
+   "source": [
+    "### import numpy as np\n",
+    "import matplotlib.pyplot as plt\n",
+    "import pandas as pd\n",
+    "import numpy as np\n",
+    "from matplotlib.colors import ListedColormap\n",
+    "from sklearn.model_selection import train_test_split\n",
+    "from sklearn.svm import SVC\n",
+    "from sklearn.tree import DecisionTreeClassifier\n",
+    "from sklearn.ensemble import RandomForestClassifier, AdaBoostClassifier\n",
+    "from sklearn.naive_bayes import GaussianNB\n",
+    "from sklearn.externals import joblib\n",
+    "\n",
+    "# These are the classifiers that permit training data with sample weights!\n",
+    "for file in files:\n",
+    "\n",
+    "    names = [\"NaiveBayes\", \"LinearSVM\", \"RBFSVM\", \"DecisionTree\",\n",
+    "         \"RandomForest\", \"AdaBoost\"]\n",
+    "\n",
+    "    classifiers = [\n",
+    "        GaussianNB(),\n",
+    "        SVC(kernel=\"linear\", C=0.025),\n",
+    "        SVC(gamma=2, C=1),\n",
+    "        DecisionTreeClassifier(max_depth=5),\n",
+    "        RandomForestClassifier(max_depth=5, n_estimators=10, max_features=1),\n",
+    "        AdaBoostClassifier()]\n",
+    "    \n",
+    "    print(\"file name :\", file)\n",
+    "    data = pd.read_csv(files[file], delimiter=r\"\\s+\").dropna()\n",
+    "    \n",
+    "#     print (data.shape[0] , data.iloc[:,0].nunique())\n",
+    "    if data.shape[0] == data.iloc[:,0].nunique():\n",
+    "        data = data.append(data)\n",
+    "#     display(data)\n",
+    "    \n",
+    "#     print(data.iloc[:,2:])\n",
+    "    \n",
+    "    # words (features) encoding\n",
+    "    from sklearn.preprocessing import OrdinalEncoder\n",
+    "    enc = OrdinalEncoder(dtype=np.int32)\n",
+    "    features = enc.fit_transform(data.iloc[:,2:])\n",
+    "#     display(enc.categories_)\n",
+    "#     display(data.iloc[:,2:],features)\n",
+    "    # target and weights\n",
+    "    target = data.iloc[:,0]\n",
+    "    weights = data.iloc[:,1].values\n",
+    "    \n",
+    "#     print(\"file name :\", file)\n",
+    "    print(\"Rules(classes) number :\",target.nunique())\n",
+    "    print(\"Words(features) number :\",features.shape[1])\n",
+    "    print(\"Records number :\",features.shape[0], end = '')\n",
+    "    display(data.iloc[:target.nunique(),:])\n",
+    "    \n",
+    "    # split to train and test\n",
+    "    X_train, X_test, y_train, y_test, w_train, w_test = \\\n",
+    "        train_test_split(features, target, weights, test_size=.5, random_state=0, stratify=target)\n",
+    "#     display(features, target, weights)\n",
+    "#     display(X_train, X_test, y_train, y_test, w_train, w_test)\n",
+    "    \n",
+    "    # train models and print their scores\n",
+    "    for name, clf in zip(names, classifiers):\n",
+    "        print(\"model :\", name, \",\", end = '')\n",
+    "        clf.fit(X=X_train, y=y_train, sample_weight=w_train)\n",
+    "        score = clf.score(X=X_test, y=y_test, sample_weight=w_test)\n",
+    "        print(\" score =\", score)\n",
+    "        \n",
+    "        # save models\n",
+    "        name+'-'+file[:-4]+'.model'\n",
+    "        filename = 'models/'+name+'-'+file[:-4]+'.model'\n",
+    "        joblib.dump(clf, filename)\n",
+    "    print(\"----------------------------------------------\\n\")\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": []
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "Python 3",
+   "language": "python",
+   "name": "python3"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 3
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython3",
+   "version": "3.7.1"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 1
+}
diff --git a/classifiers_comparison.ipynb b/classifiers_comparison.ipynb
deleted file mode 100644
index 1367409..0000000
--- a/classifiers_comparison.ipynb
+++ /dev/null
@@ -1,801 +0,0 @@
-{
- "cells": [
-  {
-   "cell_type": "code",
-   "execution_count": 1,
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "%matplotlib inline"
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "metadata": {},
-   "source": [
-    "\n",
-    "# Classifier comparison\n",
-    "\n",
-    "\n",
-    "A comparison of a several classifiers in scikit-learn on synthetic datasets.\n",
-    "The point of this example is to illustrate the nature of decision boundaries\n",
-    "of different classifiers.\n",
-    "This should be taken with a grain of salt, as the intuition conveyed by\n",
-    "these examples does not necessarily carry over to real datasets.\n",
-    "\n",
-    "Particularly in high-dimensional spaces, data can more easily be separated\n",
-    "linearly and the simplicity of classifiers such as naive Bayes and linear SVMs\n",
-    "might lead to better generalization than is achieved by other classifiers.\n",
-    "\n",
-    "The plots show training points in solid colors and testing points\n",
-    "semi-transparent. The lower right shows the classification accuracy on the test\n",
-    "set.\n",
-    "\n"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 118,
-   "metadata": {},
-   "outputs": [
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Automatically created module for IPython interactive environment\n"
-     ]
-    }
-   ],
-   "source": [
-    "print(__doc__)\n",
-    "\n",
-    "\n",
-    "# Code source: Gaël Varoquaux\n",
-    "#              Andreas Müller\n",
-    "# Modified for documentation by Jaques Grobler\n",
-    "# License: BSD 3 clause\n",
-    "\n",
-    "import numpy as np\n",
-    "import matplotlib.pyplot as plt\n",
-    "import pandas as pd\n",
-    "from matplotlib.colors import ListedColormap\n",
-    "from sklearn.model_selection import train_test_split\n",
-    "from sklearn.svm import SVC\n",
-    "from sklearn.tree import DecisionTreeClassifier\n",
-    "from sklearn.ensemble import RandomForestClassifier, AdaBoostClassifier\n",
-    "from sklearn.naive_bayes import GaussianNB\n",
-    "\n",
-    "h = .02  # step size in the mesh\n",
-    "\n",
-    "names = [\"Naive Bayes\", \"Linear SVM\", \"RBF SVM\", \"Decision Tree\",\n",
-    "         \"Random Forest\", \"AdaBoost\"]\n",
-    "\n",
-    "classifiers = [\n",
-    "    GaussianNB(),\n",
-    "    SVC(kernel=\"linear\", C=0.025),\n",
-    "    SVC(gamma=2, C=1),\n",
-    "    DecisionTreeClassifier(max_depth=5),\n",
-    "    RandomForestClassifier(max_depth=5, n_estimators=10, max_features=1),\n",
-    "    AdaBoostClassifier()]"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 73,
-   "metadata": {},
-   "outputs": [
-    {
-     "data": {
-      "text/html": [
-       "<div>\n",
-       "<style scoped>\n",
-       "    .dataframe tbody tr th:only-of-type {\n",
-       "        vertical-align: middle;\n",
-       "    }\n",
-       "\n",
-       "    .dataframe tbody tr th {\n",
-       "        vertical-align: top;\n",
-       "    }\n",
-       "\n",
-       "    .dataframe thead th {\n",
-       "        text-align: right;\n",
-       "    }\n",
-       "</style>\n",
-       "<table border=\"1\" class=\"dataframe\">\n",
-       "  <thead>\n",
-       "    <tr style=\"text-align: right;\">\n",
-       "      <th></th>\n",
-       "      <th>word1</th>\n",
-       "      <th>word2</th>\n",
-       "    </tr>\n",
-       "  </thead>\n",
-       "  <tbody>\n",
-       "    <tr>\n",
-       "      <th>0</th>\n",
-       "      <td>no</td>\n",
-       "      <td>respetar</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>1</th>\n",
-       "      <td>no</td>\n",
-       "      <td>respetar</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>2</th>\n",
-       "      <td>no</td>\n",
-       "      <td>estar</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>3</th>\n",
-       "      <td>no</td>\n",
-       "      <td>estar</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>4</th>\n",
-       "      <td>no</td>\n",
-       "      <td>caber</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>5</th>\n",
-       "      <td>no</td>\n",
-       "      <td>caber</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>6</th>\n",
-       "      <td>no</td>\n",
-       "      <td>desear</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>7</th>\n",
-       "      <td>no</td>\n",
-       "      <td>desear</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>8</th>\n",
-       "      <td>no</td>\n",
-       "      <td>ser</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>9</th>\n",
-       "      <td>no</td>\n",
-       "      <td>ser</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>10</th>\n",
-       "      <td>sólo</td>\n",
-       "      <td>reestructurar</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>11</th>\n",
-       "      <td>sólo</td>\n",
-       "      <td>reestructurar</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>12</th>\n",
-       "      <td>también</td>\n",
-       "      <td>asegurar</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>13</th>\n",
-       "      <td>también</td>\n",
-       "      <td>asegurar</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>14</th>\n",
-       "      <td>no</td>\n",
-       "      <td>deber</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>15</th>\n",
-       "      <td>no</td>\n",
-       "      <td>deber</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>16</th>\n",
-       "      <td>también</td>\n",
-       "      <td>contribuir</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>17</th>\n",
-       "      <td>también</td>\n",
-       "      <td>contribuir</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>18</th>\n",
-       "      <td>también</td>\n",
-       "      <td>ser</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>19</th>\n",
-       "      <td>también</td>\n",
-       "      <td>ser</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>20</th>\n",
-       "      <td>también</td>\n",
-       "      <td>poder</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>21</th>\n",
-       "      <td>también</td>\n",
-       "      <td>poder</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>22</th>\n",
-       "      <td>sólo</td>\n",
-       "      <td>consagrar</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>23</th>\n",
-       "      <td>sólo</td>\n",
-       "      <td>consagrar</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>24</th>\n",
-       "      <td>también</td>\n",
-       "      <td>deber</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>25</th>\n",
-       "      <td>también</td>\n",
-       "      <td>deber</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>26</th>\n",
-       "      <td>no</td>\n",
-       "      <td>ser</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>27</th>\n",
-       "      <td>no</td>\n",
-       "      <td>ser</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>28</th>\n",
-       "      <td>no</td>\n",
-       "      <td>poder</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>29</th>\n",
-       "      <td>no</td>\n",
-       "      <td>poder</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>...</th>\n",
-       "      <td>...</td>\n",
-       "      <td>...</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>23218</th>\n",
-       "      <td>también</td>\n",
-       "      <td>decir</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>23219</th>\n",
-       "      <td>también</td>\n",
-       "      <td>decir</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>23220</th>\n",
-       "      <td>no</td>\n",
-       "      <td>existir</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>23221</th>\n",
-       "      <td>no</td>\n",
-       "      <td>existir</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>23222</th>\n",
-       "      <td>no</td>\n",
-       "      <td>tener</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>23223</th>\n",
-       "      <td>no</td>\n",
-       "      <td>tener</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>23224</th>\n",
-       "      <td>ya</td>\n",
-       "      <td>encontrar</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>23225</th>\n",
-       "      <td>ya</td>\n",
-       "      <td>encontrar</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>23226</th>\n",
-       "      <td>siempre</td>\n",
-       "      <td>ser</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>23227</th>\n",
-       "      <td>siempre</td>\n",
-       "      <td>ser</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>23228</th>\n",
-       "      <td>no</td>\n",
-       "      <td>ser</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>23229</th>\n",
-       "      <td>no</td>\n",
-       "      <td>ser</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>23230</th>\n",
-       "      <td>incluso</td>\n",
-       "      <td>herir</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>23231</th>\n",
-       "      <td>incluso</td>\n",
-       "      <td>herir</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>23232</th>\n",
-       "      <td>no</td>\n",
-       "      <td>presenciar</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>23233</th>\n",
-       "      <td>no</td>\n",
-       "      <td>presenciar</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>23234</th>\n",
-       "      <td>no</td>\n",
-       "      <td>ser</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>23235</th>\n",
-       "      <td>no</td>\n",
-       "      <td>ser</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>23236</th>\n",
-       "      <td>este_año</td>\n",
-       "      <td>vencer</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>23237</th>\n",
-       "      <td>este_año</td>\n",
-       "      <td>vencer</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>23238</th>\n",
-       "      <td>no</td>\n",
-       "      <td>ser</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>23239</th>\n",
-       "      <td>no</td>\n",
-       "      <td>ser</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>23240</th>\n",
-       "      <td>sólo</td>\n",
-       "      <td>hacer</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>23241</th>\n",
-       "      <td>sólo</td>\n",
-       "      <td>hacer</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>23242</th>\n",
-       "      <td>también</td>\n",
-       "      <td>querer</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>23243</th>\n",
-       "      <td>también</td>\n",
-       "      <td>querer</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>23244</th>\n",
-       "      <td>no</td>\n",
-       "      <td>responder</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>23245</th>\n",
-       "      <td>no</td>\n",
-       "      <td>responder</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>23246</th>\n",
-       "      <td>no</td>\n",
-       "      <td>hacer</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>23247</th>\n",
-       "      <td>no</td>\n",
-       "      <td>hacer</td>\n",
-       "    </tr>\n",
-       "  </tbody>\n",
-       "</table>\n",
-       "<p>23248 rows × 2 columns</p>\n",
-       "</div>"
-      ],
-      "text/plain": [
-       "          word1          word2\n",
-       "0            no       respetar\n",
-       "1            no       respetar\n",
-       "2            no          estar\n",
-       "3            no          estar\n",
-       "4            no          caber\n",
-       "5            no          caber\n",
-       "6            no         desear\n",
-       "7            no         desear\n",
-       "8            no            ser\n",
-       "9            no            ser\n",
-       "10         sólo  reestructurar\n",
-       "11         sólo  reestructurar\n",
-       "12      también       asegurar\n",
-       "13      también       asegurar\n",
-       "14           no          deber\n",
-       "15           no          deber\n",
-       "16      también     contribuir\n",
-       "17      también     contribuir\n",
-       "18      también            ser\n",
-       "19      también            ser\n",
-       "20      también          poder\n",
-       "21      también          poder\n",
-       "22         sólo      consagrar\n",
-       "23         sólo      consagrar\n",
-       "24      también          deber\n",
-       "25      también          deber\n",
-       "26           no            ser\n",
-       "27           no            ser\n",
-       "28           no          poder\n",
-       "29           no          poder\n",
-       "...         ...            ...\n",
-       "23218   también          decir\n",
-       "23219   también          decir\n",
-       "23220        no        existir\n",
-       "23221        no        existir\n",
-       "23222        no          tener\n",
-       "23223        no          tener\n",
-       "23224        ya      encontrar\n",
-       "23225        ya      encontrar\n",
-       "23226   siempre            ser\n",
-       "23227   siempre            ser\n",
-       "23228        no            ser\n",
-       "23229        no            ser\n",
-       "23230   incluso          herir\n",
-       "23231   incluso          herir\n",
-       "23232        no     presenciar\n",
-       "23233        no     presenciar\n",
-       "23234        no            ser\n",
-       "23235        no            ser\n",
-       "23236  este_año         vencer\n",
-       "23237  este_año         vencer\n",
-       "23238        no            ser\n",
-       "23239        no            ser\n",
-       "23240      sólo          hacer\n",
-       "23241      sólo          hacer\n",
-       "23242   también         querer\n",
-       "23243   también         querer\n",
-       "23244        no      responder\n",
-       "23245        no      responder\n",
-       "23246        no          hacer\n",
-       "23247        no          hacer\n",
-       "\n",
-       "[23248 rows x 2 columns]"
-      ]
-     },
-     "metadata": {},
-     "output_type": "display_data"
-    }
-   ],
-   "source": [
-    "data = pd.read_csv(\"sklearn-nobad/75+100_37+.csv\")\n",
-    "display(data.iloc[:,2:])"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 88,
-   "metadata": {},
-   "outputs": [
-    {
-     "data": {
-      "text/plain": [
-       "array([[170, 558],\n",
-       "       [170, 558],\n",
-       "       [170, 294],\n",
-       "       ...,\n",
-       "       [170, 559],\n",
-       "       [170, 347],\n",
-       "       [170, 347]], dtype=int32)"
-      ]
-     },
-     "metadata": {},
-     "output_type": "display_data"
-    }
-   ],
-   "source": [
-    "# encode words features\n",
-    "\n",
-    "from sklearn.preprocessing import OrdinalEncoder\n",
-    "\n",
-    "enc = OrdinalEncoder(dtype=np.int32)\n",
-    "features = enc.fit_transform(data.iloc[:,2:])\n",
-    "display(features)\n",
-    "\n",
-    "# enc.categories_\n",
-    "\n",
-    "# It gives error if unseen word!\n",
-    "# display(enc.transform([['a_ciencia_cierta', 'añadir']]))"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 105,
-   "metadata": {},
-   "outputs": [
-    {
-     "data": {
-      "text/plain": [
-       "0        0\n",
-       "1        1\n",
-       "2        0\n",
-       "3        1\n",
-       "4        0\n",
-       "5        1\n",
-       "6        0\n",
-       "7        1\n",
-       "8        0\n",
-       "9        1\n",
-       "10       0\n",
-       "11       1\n",
-       "12       0\n",
-       "13       1\n",
-       "14       0\n",
-       "15       1\n",
-       "16       0\n",
-       "17       1\n",
-       "18       0\n",
-       "19       1\n",
-       "20       0\n",
-       "21       1\n",
-       "22       0\n",
-       "23       1\n",
-       "24       0\n",
-       "25       1\n",
-       "26       0\n",
-       "27       1\n",
-       "28       0\n",
-       "29       1\n",
-       "        ..\n",
-       "23218    0\n",
-       "23219    1\n",
-       "23220    0\n",
-       "23221    1\n",
-       "23222    0\n",
-       "23223    1\n",
-       "23224    0\n",
-       "23225    1\n",
-       "23226    0\n",
-       "23227    1\n",
-       "23228    0\n",
-       "23229    1\n",
-       "23230    0\n",
-       "23231    1\n",
-       "23232    0\n",
-       "23233    1\n",
-       "23234    0\n",
-       "23235    1\n",
-       "23236    0\n",
-       "23237    1\n",
-       "23238    0\n",
-       "23239    1\n",
-       "23240    0\n",
-       "23241    1\n",
-       "23242    0\n",
-       "23243    1\n",
-       "23244    0\n",
-       "23245    1\n",
-       "23246    0\n",
-       "23247    1\n",
-       "Name: rule, Length: 23248, dtype: int64"
-      ]
-     },
-     "metadata": {},
-     "output_type": "display_data"
-    },
-    {
-     "data": {
-      "text/plain": [
-       "array([0.5158550000000001, 0.484145, 0.50555, ..., 0.49348400000000003,\n",
-       "       0.45832799999999996, 0.541672], dtype=object)"
-      ]
-     },
-     "metadata": {},
-     "output_type": "display_data"
-    }
-   ],
-   "source": [
-    "# target and sample weights\n",
-    "target = data.iloc[:,0]\n",
-    "display(target)\n",
-    "\n",
-    "weights = data.values[:,1]\n",
-    "display(weights)"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 106,
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "X_train, X_test, y_train, y_test, w_train, w_test = \\\n",
-    "        train_test_split(features, target, weights, test_size=.3)"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 119,
-   "metadata": {},
-   "outputs": [
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "model : Naive Bayes , score = 0.49680499419261925\n",
-      "model : Linear SVM , score = 0.5016183944944038\n",
-      "model : RBF SVM , score = 0.3670043827474011\n",
-      "model : Decision Tree , score = 0.4847892356844361\n",
-      "model : Random Forest , score = 0.47018256686956\n",
-      "model : AdaBoost , score = 0.46697713150721004\n"
-     ]
-    }
-   ],
-   "source": [
-    "for name, clf in zip(names, classifiers):\n",
-    "    print(\"model :\", name, \",\", end = '')\n",
-    "    clf.fit(X=X_train, y=y_train, sample_weight=w_train)\n",
-    "    score = clf.score(X=X_test, y=y_test, sample_weight=w_test)\n",
-    "    print(\" score =\", score)"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 2,
-   "metadata": {},
-   "outputs": [
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Automatically created module for IPython interactive environment\n"
-     ]
-    },
-    {
-     "data": {
-      "image/png": "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\n",
-      "text/plain": [
-       "<Figure size 1944x648 with 33 Axes>"
-      ]
-     },
-     "metadata": {},
-     "output_type": "display_data"
-    }
-   ],
-   "source": [
-    "X, y = make_classification(n_features=2, n_redundant=0, n_informative=2,\n",
-    "                           random_state=1, n_clusters_per_class=1)\n",
-    "rng = np.random.RandomState(2)\n",
-    "X += 2 * rng.uniform(size=X.shape)\n",
-    "linearly_separable = (X, y)\n",
-    "\n",
-    "datasets = [make_moons(noise=0.3, random_state=0),\n",
-    "            make_circles(noise=0.2, factor=0.5, random_state=1),\n",
-    "            linearly_separable\n",
-    "            ]\n",
-    "\n",
-    "figure = plt.figure(figsize=(27, 9))\n",
-    "i = 1\n",
-    "# iterate over datasets\n",
-    "for ds_cnt, ds in enumerate(datasets):\n",
-    "    # preprocess dataset, split into training and test part\n",
-    "    X, y = ds\n",
-    "    X = StandardScaler().fit_transform(X)\n",
-    "    X_train, X_test, y_train, y_test = \\\n",
-    "        train_test_split(X, y, test_size=.4, random_state=42)\n",
-    "\n",
-    "    x_min, x_max = X[:, 0].min() - .5, X[:, 0].max() + .5\n",
-    "    y_min, y_max = X[:, 1].min() - .5, X[:, 1].max() + .5\n",
-    "    xx, yy = np.meshgrid(np.arange(x_min, x_max, h),\n",
-    "                         np.arange(y_min, y_max, h))\n",
-    "\n",
-    "    # just plot the dataset first\n",
-    "    cm = plt.cm.RdBu\n",
-    "    cm_bright = ListedColormap(['#FF0000', '#0000FF'])\n",
-    "    ax = plt.subplot(len(datasets), len(classifiers) + 1, i)\n",
-    "    if ds_cnt == 0:\n",
-    "        ax.set_title(\"Input data\")\n",
-    "    # Plot the training points\n",
-    "    ax.scatter(X_train[:, 0], X_train[:, 1], c=y_train, cmap=cm_bright,\n",
-    "               edgecolors='k')\n",
-    "    # Plot the testing points\n",
-    "    ax.scatter(X_test[:, 0], X_test[:, 1], c=y_test, cmap=cm_bright, alpha=0.6,\n",
-    "               edgecolors='k')\n",
-    "    ax.set_xlim(xx.min(), xx.max())\n",
-    "    ax.set_ylim(yy.min(), yy.max())\n",
-    "    ax.set_xticks(())\n",
-    "    ax.set_yticks(())\n",
-    "    i += 1\n",
-    "\n",
-    "    # iterate over classifiers\n",
-    "    for name, clf in zip(names, classifiers):\n",
-    "        ax = plt.subplot(len(datasets), len(classifiers) + 1, i)\n",
-    "        clf.fit(X_train, y_train)\n",
-    "        score = clf.score(X_test, y_test)\n",
-    "\n",
-    "        # Plot the decision boundary. For that, we will assign a color to each\n",
-    "        # point in the mesh [x_min, x_max]x[y_min, y_max].\n",
-    "        if hasattr(clf, \"decision_function\"):\n",
-    "            Z = clf.decision_function(np.c_[xx.ravel(), yy.ravel()])\n",
-    "        else:\n",
-    "            Z = clf.predict_proba(np.c_[xx.ravel(), yy.ravel()])[:, 1]\n",
-    "\n",
-    "        # Put the result into a color plot\n",
-    "        Z = Z.reshape(xx.shape)\n",
-    "        ax.contourf(xx, yy, Z, cmap=cm, alpha=.8)\n",
-    "\n",
-    "        # Plot the training points\n",
-    "        ax.scatter(X_train[:, 0], X_train[:, 1], c=y_train, cmap=cm_bright,\n",
-    "                   edgecolors='k')\n",
-    "        # Plot the testing points\n",
-    "        ax.scatter(X_test[:, 0], X_test[:, 1], c=y_test, cmap=cm_bright,\n",
-    "                   edgecolors='k', alpha=0.6)\n",
-    "\n",
-    "        ax.set_xlim(xx.min(), xx.max())\n",
-    "        ax.set_ylim(yy.min(), yy.max())\n",
-    "        ax.set_xticks(())\n",
-    "        ax.set_yticks(())\n",
-    "        if ds_cnt == 0:\n",
-    "            ax.set_title(name)\n",
-    "        ax.text(xx.max() - .3, yy.min() + .3, ('%.2f' % score).lstrip('0'),\n",
-    "                size=15, horizontalalignment='right')\n",
-    "        i += 1\n",
-    "\n",
-    "plt.tight_layout()\n",
-    "plt.show()"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": null,
-   "metadata": {},
-   "outputs": [],
-   "source": []
-  }
- ],
- "metadata": {
-  "kernelspec": {
-   "display_name": "Python 3",
-   "language": "python",
-   "name": "python3"
-  },
-  "language_info": {
-   "codemirror_mode": {
-    "name": "ipython",
-    "version": 3
-   },
-   "file_extension": ".py",
-   "mimetype": "text/x-python",
-   "name": "python",
-   "nbconvert_exporter": "python",
-   "pygments_lexer": "ipython3",
-   "version": "3.7.1"
-  }
- },
- "nbformat": 4,
- "nbformat_minor": 1
-}