4. Análisis léxico II.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "![agents](images/header.jpg)\n",
    "# Análisis léxico: Colocaciones y $N$-Gramas\n",
    "### Ramón Soto C. [(rsotoc@moviquest.com)](mailto:rsotoc@moviquest.com/)\n",
    "[ver en nbviewer](http://nbviewer.ipython.org/github/rsotoc/nlp/blob/master/Introducción.ipynb)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n",
    "\n",
    "## Modelo de N-Gramas\n",
    "\n",
    "Una característica de los lenguajes naturales es que las frases que los componen no tienen una distribución uniforme. Por el contrario, existen construcciones que son más comunes que otras. De esta manera, aunque las frases $f_1 = \\textrm{\"el perro camina\"}$ y $f_2 = \\textrm{\"el perro vuela\"}$ son ambas correctas sintácticamente, es más probable encontrar la frase $f_1$ que la frase $f_2$ en un texto arbitrario. \n",
    "\n",
    "En la película \"*Take The Money And Run*\", Virgil Starkwell (Woody Allen) intenta asaltar un banco y entrega al cajero una nota con el mensaje \"*Please put fifty thousand dollars into this bag and act natural as I am pointing a gun at you*\" que es leída por los empleados del banco como \"*Please put fifty thousand dollars into this bag and ABT natural as I am pointing a GUB at you*\". \n",
    "\n",
    "[![](images/i_have_a_gub.jpg)](https://www.youtube.com/watch?v=pEm0zi8QrpA)\n",
    "\n",
    "Sin embargo, es obvio que la frase \"*I am pointing a GUN at you*\" es más probable que la frase \"*I am pointing a GUB at you*\", por lo que en la vida real no nos cuesta trabajo reconocer la frase correcta. Para modelar esta capacidad de predecir la ocurrencia de una palabra en una frase se utilizan **Modelos de lenguajes** que asignan probabilidades a las secuencias de palabras que pueden conformar un texto. \n",
    "\n",
    "El modelo más simple es el **Modelo de N-Gramas**\". Este modelo asume que la probabilidad de ocurrencia de una palabra está determinada por las palabras recientes; lo que se conoce como la **suposición de Markov**. De manera que para el cálculo de estas probabilidades basta contabilizar la ocurrencia de secuencias de palabras de longitud definida. Un **$N$-Grama** es una secuencia de $N$ palabras. Así, por ejemplo, un 2-grama (o bigrama) es una secuencia de 2 palabras, como \"*el hombre*\", \"*hombre camina*\", \"*camina en*\", \"*en el*\", \"*el parque*\". Un 3-grama (trigrama) es una secuencia de tres palabras, como \"*el hombre camina*\", \"*hombre camina en*\", \"*camina en el*\", \"*en el parque*\". "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Bigramas:\n",
      " Counter({('.', 'el'): 2, ('el', 'perro'): 2, ('en', 'el'): 2, ('el', 'parque'): 2, ('los', 'amigos'): 1, ('amigos', 'toman'): 1, ('toman', 'cafe'): 1, ('cafe', '.'): 1, ('perro', 'duerme'): 1, ('duerme', 'en'): 1, ('parque', '.'): 1, ('el', 'hombre'): 1, ('hombre', 'con'): 1, ('con', 'el'): 1, ('perro', 'camina'): 1, ('camina', 'en'): 1, ('parque', 'con'): 1, ('con', 'amigos'): 1, ('amigos', '.'): 1})\n",
      "\n",
      "Trigramas:\n",
      " Counter({('en', 'el', 'parque'): 2, ('los', 'amigos', 'toman'): 1, ('amigos', 'toman', 'cafe'): 1, ('toman', 'cafe', '.'): 1, ('cafe', '.', 'el'): 1, ('.', 'el', 'perro'): 1, ('el', 'perro', 'duerme'): 1, ('perro', 'duerme', 'en'): 1, ('duerme', 'en', 'el'): 1, ('el', 'parque', '.'): 1, ('parque', '.', 'el'): 1, ('.', 'el', 'hombre'): 1, ('el', 'hombre', 'con'): 1, ('hombre', 'con', 'el'): 1, ('con', 'el', 'perro'): 1, ('el', 'perro', 'camina'): 1, ('perro', 'camina', 'en'): 1, ('camina', 'en', 'el'): 1, ('el', 'parque', 'con'): 1, ('parque', 'con', 'amigos'): 1, ('con', 'amigos', '.'): 1})\n",
      "\n",
      "Tetragramas:\n",
      " Counter({('los', 'amigos', 'toman', 'cafe'): 1, ('amigos', 'toman', 'cafe', '.'): 1, ('toman', 'cafe', '.', 'el'): 1, ('cafe', '.', 'el', 'perro'): 1, ('.', 'el', 'perro', 'duerme'): 1, ('el', 'perro', 'duerme', 'en'): 1, ('perro', 'duerme', 'en', 'el'): 1, ('duerme', 'en', 'el', 'parque'): 1, ('en', 'el', 'parque', '.'): 1, ('el', 'parque', '.', 'el'): 1, ('parque', '.', 'el', 'hombre'): 1, ('.', 'el', 'hombre', 'con'): 1, ('el', 'hombre', 'con', 'el'): 1, ('hombre', 'con', 'el', 'perro'): 1, ('con', 'el', 'perro', 'camina'): 1, ('el', 'perro', 'camina', 'en'): 1, ('perro', 'camina', 'en', 'el'): 1, ('camina', 'en', 'el', 'parque'): 1, ('en', 'el', 'parque', 'con'): 1, ('el', 'parque', 'con', 'amigos'): 1, ('parque', 'con', 'amigos', '.'): 1})\n"
     ]
    }
   ],
   "source": [
    "import nltk\n",
    "from nltk import word_tokenize\n",
    "from nltk.util import ngrams\n",
    "from collections import Counter\n",
    "\n",
    "text = \"los amigos toman cafe. el perro duerme en el parque. el hombre con el perro \\\n",
    "camina en el parque con amigos.\"\n",
    "\n",
    "token = nltk.word_tokenize(text)\n",
    "bigrams = ngrams(token,2)\n",
    "trigrams = ngrams(token,3)\n",
    "tetragrams = ngrams(token,4)\n",
    "\n",
    "bigrams_list = list(bigrams)\n",
    "counter_bigrams = Counter(bigrams_list)\n",
    "print (\"Bigramas:\\n\", counter_bigrams)\n",
    "\n",
    "trigrams_list = list(trigrams)\n",
    "counter_trigrams = Counter(trigrams_list)\n",
    "print (\"\\nTrigramas:\\n\", counter_trigrams)\n",
    "\n",
    "tetragrams_list = list(tetragrams)\n",
    "counter_tetragrams = Counter(tetragrams_list)\n",
    "print (\"\\nTetragramas:\\n\", counter_tetragrams)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### $N$-Gramas en la base de datos de personajes de comics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "from IPython.display import display\n",
    "import pandas as pd\n",
    "pd.options.display.max_colwidth = 150 "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>description</th>\n",
       "      <th>main_words</th>\n",
       "      <th>name</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>mazing man is the title character of a comic book series created by bob rozakis and stephen destefano and published by dc comics the series ran fo...</td>\n",
       "      <td>[man, title, character, comic, book, series, created, bob, rozakis, stephen, published, dc, comics, series, ran, twelve, issues, additional, speci...</td>\n",
       "      <td>'Mazing Man</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                                                                                                             description  \\\n",
       "0  mazing man is the title character of a comic book series created by bob rozakis and stephen destefano and published by dc comics the series ran fo...   \n",
       "\n",
       "                                                                                                                                              main_words  \\\n",
       "0  [man, title, character, comic, book, series, created, bob, rozakis, stephen, published, dc, comics, series, ran, twelve, issues, additional, speci...   \n",
       "\n",
       "          name  \n",
       "0  'Mazing Man  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import json\n",
    "\n",
    "file = 'Data Sets/Comics/clean_comics.json'\n",
    "with open(file) as comics_file:\n",
    "    dict_comics = json.load(comics_file)\n",
    "\n",
    "comicsDf = pd.DataFrame.from_dict(dict_comics)\n",
    "\n",
    "display(comicsDf.head(1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>name</th>\n",
       "      <th>description</th>\n",
       "      <th>main_words</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>'Mazing Man</td>\n",
       "      <td>mazing man is the title character of a comic book series created by bob rozakis and stephen destefano and published by dc comics the series ran fo...</td>\n",
       "      <td>[man, title, character, comic, book, series, created, bob, rozakis, stephen, published, dc, comics, series, ran, twelve, issues, additional, speci...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          name  \\\n",
       "0  'Mazing Man   \n",
       "\n",
       "                                                                                                                                             description  \\\n",
       "0  mazing man is the title character of a comic book series created by bob rozakis and stephen destefano and published by dc comics the series ran fo...   \n",
       "\n",
       "                                                                                                                                              main_words  \n",
       "0  [man, title, character, comic, book, series, created, bob, rozakis, stephen, published, dc, comics, series, ran, twelve, issues, additional, speci...  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#Sólo para reducir mi OCD\n",
    "comicsDf = comicsDf.reindex_axis(['name',\"description\", \"main_words\"], axis=1)\n",
    "display(comicsDf.head(1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/rsotoc/anaconda3/lib/python3.6/site-packages/ipykernel/__main__.py:1: DeprecationWarning: generator 'ngrams' raised StopIteration\n",
      "  if __name__ == '__main__':\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>name</th>\n",
       "      <th>description</th>\n",
       "      <th>main_words</th>\n",
       "      <th>bigrams</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>'Mazing Man</td>\n",
       "      <td>mazing man is the title character of a comic book series created by bob rozakis and stephen destefano and published by dc comics the series ran fo...</td>\n",
       "      <td>[man, title, character, comic, book, series, created, bob, rozakis, stephen, published, dc, comics, series, ran, twelve, issues, additional, speci...</td>\n",
       "      <td>[(mazing, man), (man, is), (is, the), (the, title), (title, character), (character, of), (of, a), (a, comic), (comic, book), (book, series), (seri...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>711 (Quality Comics)</td>\n",
       "      <td>is a fictional superhero from the golden age of comics he was created by george brenner and published by quality comics first appeared in police c...</td>\n",
       "      <td>[fictional, superhero, golden, age, comics, created, george, published, quality, comics, first, appeared, police, comics, august, lasted, january,...</td>\n",
       "      <td>[(is, a), (a, fictional), (fictional, superhero), (superhero, from), (from, the), (the, golden), (golden, age), (age, of), (of, comics), (comics, ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Abigail Brand</td>\n",
       "      <td>special agent special agent abigail brand is a fictional character appearing in american comic book s published by marvel comics publication histo...</td>\n",
       "      <td>[special, agent, special, agent, abigail, brand, fictional, character, appearing, american, comic, book, published, marvel, comics, publication, h...</td>\n",
       "      <td>[(special, agent), (agent, special), (special, agent), (agent, abigail), (abigail, brand), (brand, is), (is, a), (a, fictional), (fictional, chara...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Abin Sur</td>\n",
       "      <td>abin sur is a fictional character and a superhero from the dc comics dc universe he was a member of the green lantern corps and is best known as t...</td>\n",
       "      <td>[abin, sur, fictional, character, superhero, dc, comics, dc, universe, member, green, lantern, corps, best, known, predecessor, green, lantern, ha...</td>\n",
       "      <td>[(abin, sur), (sur, is), (is, a), (a, fictional), (fictional, character), (character, and), (and, a), (a, superhero), (superhero, from), (from, th...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Abner Jenkins</td>\n",
       "      <td>abner ronald jenkins formerly known as the beetle comics beetle mach iv mach v mach vii and currently known as mach x and is a fictional character...</td>\n",
       "      <td>[abner, ronald, jenkins, formerly, known, beetle, comics, beetle, mach, mach, mach, vii, currently, known, mach, x, fictional, character, appearin...</td>\n",
       "      <td>[(abner, ronald), (ronald, jenkins), (jenkins, formerly), (formerly, known), (known, as), (as, the), (the, beetle), (beetle, comics), (comics, bee...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   name  \\\n",
       "0           'Mazing Man   \n",
       "1  711 (Quality Comics)   \n",
       "2         Abigail Brand   \n",
       "3              Abin Sur   \n",
       "4         Abner Jenkins   \n",
       "\n",
       "                                                                                                                                             description  \\\n",
       "0  mazing man is the title character of a comic book series created by bob rozakis and stephen destefano and published by dc comics the series ran fo...   \n",
       "1  is a fictional superhero from the golden age of comics he was created by george brenner and published by quality comics first appeared in police c...   \n",
       "2  special agent special agent abigail brand is a fictional character appearing in american comic book s published by marvel comics publication histo...   \n",
       "3  abin sur is a fictional character and a superhero from the dc comics dc universe he was a member of the green lantern corps and is best known as t...   \n",
       "4  abner ronald jenkins formerly known as the beetle comics beetle mach iv mach v mach vii and currently known as mach x and is a fictional character...   \n",
       "\n",
       "                                                                                                                                              main_words  \\\n",
       "0  [man, title, character, comic, book, series, created, bob, rozakis, stephen, published, dc, comics, series, ran, twelve, issues, additional, speci...   \n",
       "1  [fictional, superhero, golden, age, comics, created, george, published, quality, comics, first, appeared, police, comics, august, lasted, january,...   \n",
       "2  [special, agent, special, agent, abigail, brand, fictional, character, appearing, american, comic, book, published, marvel, comics, publication, h...   \n",
       "3  [abin, sur, fictional, character, superhero, dc, comics, dc, universe, member, green, lantern, corps, best, known, predecessor, green, lantern, ha...   \n",
       "4  [abner, ronald, jenkins, formerly, known, beetle, comics, beetle, mach, mach, mach, vii, currently, known, mach, x, fictional, character, appearin...   \n",
       "\n",
       "                                                                                                                                                 bigrams  \n",
       "0  [(mazing, man), (man, is), (is, the), (the, title), (title, character), (character, of), (of, a), (a, comic), (comic, book), (book, series), (seri...  \n",
       "1  [(is, a), (a, fictional), (fictional, superhero), (superhero, from), (from, the), (the, golden), (golden, age), (age, of), (of, comics), (comics, ...  \n",
       "2  [(special, agent), (agent, special), (special, agent), (agent, abigail), (abigail, brand), (brand, is), (is, a), (a, fictional), (fictional, chara...  \n",
       "3  [(abin, sur), (sur, is), (is, a), (a, fictional), (fictional, character), (character, and), (and, a), (a, superhero), (superhero, from), (from, th...  \n",
       "4  [(abner, ronald), (ronald, jenkins), (jenkins, formerly), (formerly, known), (known, as), (as, the), (the, beetle), (beetle, comics), (comics, bee...  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Cantidad de bigramas en el corpus:  3658794\n",
      "\n",
      "Bigramas más populares:\n",
      " [(('of', 'the'), 28477), (('in', 'the'), 23622), (('x', 'men'), 9076), (('to', 'the'), 8773), (('marvel', 'comics'), 8651), (('and', 'the'), 8417), (('as', 'a'), 7959), (('with', 'the'), 7353), (('dc', 'comics'), 6006), (('by', 'the'), 5953), (('he', 'is'), 5579), (('in', 'a'), 5472), (('to', 'be'), 5203), (('spider', 'man'), 5192), (('is', 'a'), 4973), (('from', 'the'), 4648), (('on', 'the'), 4431), (('as', 'the'), 4374), (('of', 'his'), 4259), (('created', 'by'), 4168), (('he', 'was'), 4066), (('at', 'the'), 3881), (('category', 'fictional'), 3822), (('the', 'new'), 3721), (('for', 'the'), 3635), (('one', 'of'), 3621), (('during', 'the'), 3605), (('that', 'he'), 3455), (('the', 'character'), 3450), (('comic', 'book'), 3373), (('justice', 'league'), 3340), (('of', 'a'), 3300), (('the', 'x'), 3262), (('she', 'is'), 3209), (('the', 'team'), 2970), (('comics', 'characters'), 2937), (('appears', 'in'), 2889), (('appeared', 'in'), 2876), (('member', 'of'), 2856), (('version', 'of'), 2625), (('with', 'a'), 2598), (('him', 'to'), 2577), (('that', 'the'), 2569), (('fictional', 'character'), 2562), (('it', 'is'), 2555), (('able', 'to'), 2495), (('green', 'lantern'), 2418), (('and', 'his'), 2404), (('the', 'first'), 2383), (('category', 'marvel'), 2335)]\n"
     ]
    }
   ],
   "source": [
    "comicsDf[\"bigrams\"] = list(map(lambda row: list(ngrams(word_tokenize(row),2)), \n",
    "                                   comicsDf.description))\n",
    "display(comicsDf.head())\n",
    "\n",
    "comics_bigrams = []\n",
    "for row in comicsDf.bigrams:\n",
    "    comics_bigrams.extend(row)\n",
    "most_common_comics_bigrams = nltk.FreqDist(comics_bigrams)\n",
    "\n",
    "print(\"Cantidad de bigramas en el corpus: \", most_common_comics_bigrams.N())\n",
    "print(\"\\nBigramas más populares:\\n\", most_common_comics_bigrams.most_common(50))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Colocaciones\n",
    "\n",
    "En lexicología, una colocación es un término que designa combinaciones de unidades léxicas  que se distinguen por una frecuencia que resulta inesperadamente alta. En los bigramas anteriores destacan los siguientes:\n",
    "\n",
    "<code>[... (('x', 'men'), 8457), ... (('marvel', 'comics'), 6206), ... (('spider', 'man'), 4942), ... (('dc', 'comics'), 3906), ... (('comic', 'book'), 3280), ... (('justice', 'league'), 2991), ... (('green', 'lantern'), 2288), ...]\n",
    "</code>\n",
    "\n",
    "Estos términos permiten hacer un análisis más inteligente del texto. Identificamos otras colocaciones importantes:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/rsotoc/anaconda3/lib/python3.6/site-packages/ipykernel/__main__.py:2: DeprecationWarning: generator 'ngrams' raised StopIteration\n",
      "  from ipykernel import kernelapp as app\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>name</th>\n",
       "      <th>description</th>\n",
       "      <th>main_words</th>\n",
       "      <th>bigrams</th>\n",
       "      <th>clean_bigrams</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>'Mazing Man</td>\n",
       "      <td>mazing man is the title character of a comic book series created by bob rozakis and stephen destefano and published by dc comics the series ran fo...</td>\n",
       "      <td>[man, title, character, comic, book, series, created, bob, rozakis, stephen, published, dc, comics, series, ran, twelve, issues, additional, speci...</td>\n",
       "      <td>[(mazing, man), (man, is), (is, the), (the, title), (title, character), (character, of), (of, a), (a, comic), (comic, book), (book, series), (seri...</td>\n",
       "      <td>[(title, character), (comic, book), (book, series), (series, created), (bob, rozakis), (dc, comics), (series, ran), (twelve, issues), (additional,...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>711 (Quality Comics)</td>\n",
       "      <td>is a fictional superhero from the golden age of comics he was created by george brenner and published by quality comics first appeared in police c...</td>\n",
       "      <td>[fictional, superhero, golden, age, comics, created, george, published, quality, comics, first, appeared, police, comics, august, lasted, january,...</td>\n",
       "      <td>[(is, a), (a, fictional), (fictional, superhero), (superhero, from), (from, the), (the, golden), (golden, age), (age, of), (of, comics), (comics, ...</td>\n",
       "      <td>[(fictional, superhero), (golden, age), (quality, comics), (comics, first), (first, appeared), (police, comics), (comics, august), (killed, fictio...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Abigail Brand</td>\n",
       "      <td>special agent special agent abigail brand is a fictional character appearing in american comic book s published by marvel comics publication histo...</td>\n",
       "      <td>[special, agent, special, agent, abigail, brand, fictional, character, appearing, american, comic, book, published, marvel, comics, publication, h...</td>\n",
       "      <td>[(special, agent), (agent, special), (special, agent), (agent, abigail), (abigail, brand), (brand, is), (is, a), (a, fictional), (fictional, chara...</td>\n",
       "      <td>[(special, agent), (agent, special), (special, agent), (agent, abigail), (abigail, brand), (fictional, character), (character, appearing), (americ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Abin Sur</td>\n",
       "      <td>abin sur is a fictional character and a superhero from the dc comics dc universe he was a member of the green lantern corps and is best known as t...</td>\n",
       "      <td>[abin, sur, fictional, character, superhero, dc, comics, dc, universe, member, green, lantern, corps, best, known, predecessor, green, lantern, ha...</td>\n",
       "      <td>[(abin, sur), (sur, is), (is, a), (a, fictional), (fictional, character), (character, and), (and, a), (a, superhero), (superhero, from), (from, th...</td>\n",
       "      <td>[(abin, sur), (fictional, character), (dc, comics), (comics, dc), (dc, universe), (green, lantern), (lantern, corps), (best, known), (green, lante...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Abner Jenkins</td>\n",
       "      <td>abner ronald jenkins formerly known as the beetle comics beetle mach iv mach v mach vii and currently known as mach x and is a fictional character...</td>\n",
       "      <td>[abner, ronald, jenkins, formerly, known, beetle, comics, beetle, mach, mach, mach, vii, currently, known, mach, x, fictional, character, appearin...</td>\n",
       "      <td>[(abner, ronald), (ronald, jenkins), (jenkins, formerly), (formerly, known), (known, as), (as, the), (the, beetle), (beetle, comics), (comics, bee...</td>\n",
       "      <td>[(abner, ronald), (ronald, jenkins), (jenkins, formerly), (formerly, known), (beetle, comics), (comics, beetle), (beetle, mach), (mach, vii), (cur...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   name  \\\n",
       "0           'Mazing Man   \n",
       "1  711 (Quality Comics)   \n",
       "2         Abigail Brand   \n",
       "3              Abin Sur   \n",
       "4         Abner Jenkins   \n",
       "\n",
       "                                                                                                                                             description  \\\n",
       "0  mazing man is the title character of a comic book series created by bob rozakis and stephen destefano and published by dc comics the series ran fo...   \n",
       "1  is a fictional superhero from the golden age of comics he was created by george brenner and published by quality comics first appeared in police c...   \n",
       "2  special agent special agent abigail brand is a fictional character appearing in american comic book s published by marvel comics publication histo...   \n",
       "3  abin sur is a fictional character and a superhero from the dc comics dc universe he was a member of the green lantern corps and is best known as t...   \n",
       "4  abner ronald jenkins formerly known as the beetle comics beetle mach iv mach v mach vii and currently known as mach x and is a fictional character...   \n",
       "\n",
       "                                                                                                                                              main_words  \\\n",
       "0  [man, title, character, comic, book, series, created, bob, rozakis, stephen, published, dc, comics, series, ran, twelve, issues, additional, speci...   \n",
       "1  [fictional, superhero, golden, age, comics, created, george, published, quality, comics, first, appeared, police, comics, august, lasted, january,...   \n",
       "2  [special, agent, special, agent, abigail, brand, fictional, character, appearing, american, comic, book, published, marvel, comics, publication, h...   \n",
       "3  [abin, sur, fictional, character, superhero, dc, comics, dc, universe, member, green, lantern, corps, best, known, predecessor, green, lantern, ha...   \n",
       "4  [abner, ronald, jenkins, formerly, known, beetle, comics, beetle, mach, mach, mach, vii, currently, known, mach, x, fictional, character, appearin...   \n",
       "\n",
       "                                                                                                                                                 bigrams  \\\n",
       "0  [(mazing, man), (man, is), (is, the), (the, title), (title, character), (character, of), (of, a), (a, comic), (comic, book), (book, series), (seri...   \n",
       "1  [(is, a), (a, fictional), (fictional, superhero), (superhero, from), (from, the), (the, golden), (golden, age), (age, of), (of, comics), (comics, ...   \n",
       "2  [(special, agent), (agent, special), (special, agent), (agent, abigail), (abigail, brand), (brand, is), (is, a), (a, fictional), (fictional, chara...   \n",
       "3  [(abin, sur), (sur, is), (is, a), (a, fictional), (fictional, character), (character, and), (and, a), (a, superhero), (superhero, from), (from, th...   \n",
       "4  [(abner, ronald), (ronald, jenkins), (jenkins, formerly), (formerly, known), (known, as), (as, the), (the, beetle), (beetle, comics), (comics, bee...   \n",
       "\n",
       "                                                                                                                                           clean_bigrams  \n",
       "0  [(title, character), (comic, book), (book, series), (series, created), (bob, rozakis), (dc, comics), (series, ran), (twelve, issues), (additional,...  \n",
       "1  [(fictional, superhero), (golden, age), (quality, comics), (comics, first), (first, appeared), (police, comics), (comics, august), (killed, fictio...  \n",
       "2  [(special, agent), (agent, special), (special, agent), (agent, abigail), (abigail, brand), (fictional, character), (character, appearing), (americ...  \n",
       "3  [(abin, sur), (fictional, character), (dc, comics), (comics, dc), (dc, universe), (green, lantern), (lantern, corps), (best, known), (green, lante...  \n",
       "4  [(abner, ronald), (ronald, jenkins), (jenkins, formerly), (formerly, known), (beetle, comics), (comics, beetle), (beetle, mach), (mach, vii), (cur...  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Cantidad de bigramas en el corpus:  1080272\n",
      "\n",
      "Bigramas más populares:\n",
      " [(('x', 'men'), 9076), (('marvel', 'comics'), 8656), (('dc', 'comics'), 6009), (('spider', 'man'), 5192), (('category', 'fictional'), 3822), (('comic', 'book'), 3374), (('justice', 'league'), 3340), (('comics', 'characters'), 2944), (('fictional', 'character'), 2562), (('green', 'lantern'), 2419), (('category', 'marvel'), 2335), (('category', 'comics'), 2223), (('characters', 'created'), 2146), (('category', 'characters'), 2137), (('captain', 'america'), 2064), (('category', 'dc'), 1810), (('characters', 'introduced'), 1793), (('superheroes', 'category'), 1749), (('comics', 'superheroes'), 1649), (('tv', 'series'), 1566), (('character', 'biography'), 1537), (('iron', 'man'), 1522), (('first', 'appeared'), 1474), (('fantastic', 'four'), 1425), (('teen', 'titans'), 1400), (('uncanny', 'x'), 1387), (('external', 'links'), 1377), (('wonder', 'woman'), 1300), (('super', 'heroes'), 1160), (('new', 'york'), 1097), (('publication', 'history'), 1045), (('captain', 'marvel'), 1016), (('marvel', 'universe'), 1004), (('civil', 'war'), 963), (('new', 'mutants'), 908), (('superhuman', 'strength'), 903), (('x', 'factor'), 889), (('x', 'force'), 869), (('fictional', 'characters'), 856), (('dc', 'universe'), 847), (('lantern', 'corps'), 839), (('golden', 'age'), 824), (('playable', 'character'), 792), (('american', 'comic'), 784), (('justice', 'society'), 781), (('ghost', 'rider'), 780), (('united', 'states'), 762), (('amazing', 'spider'), 758), (('world', 'war'), 712), (('green', 'arrow'), 712)]\n"
     ]
    }
   ],
   "source": [
    "comicsDf[\"clean_bigrams\"] = list(map(lambda words, bigrams: \n",
    "                                        [b for b in list(ngrams(words, 2))\n",
    "                                        if b in bigrams], \n",
    "                                        comicsDf.main_words, comicsDf.bigrams))\n",
    "display(comicsDf.head())\n",
    "\n",
    "clean_bigrams = []\n",
    "for row in comicsDf.clean_bigrams:\n",
    "    clean_bigrams.extend(row)\n",
    "common_clean_bigrams = nltk.FreqDist(clean_bigrams)\n",
    "\n",
    "print(\"Cantidad de bigramas en el corpus: \", common_clean_bigrams.N())\n",
    "print(\"\\nBigramas más populares:\\n\", common_clean_bigrams.most_common(50))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "En estos resultados observamos que todos los bigramas presentados tienen significado útil y algunos como <code>'x', 'men'</code>, <code>'spider', 'man'</code> y muchos otros, pierden sentido si se les separa. En estos casos, puede ser conveniente reemplazar el bigrama por un token: <code>'x_men'</code>, <code>'spider_man'</code>, etc. A continuación, generamos las colocaciones para todos los documentos (*all_collocations*) y analizamos sus estadísticas en el corpus."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>name</th>\n",
       "      <th>description</th>\n",
       "      <th>main_words</th>\n",
       "      <th>bigrams</th>\n",
       "      <th>clean_bigrams</th>\n",
       "      <th>all_collocations</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>'Mazing Man</td>\n",
       "      <td>mazing man is the title character of a comic book series created by bob rozakis and stephen destefano and published by dc comics the series ran fo...</td>\n",
       "      <td>[man, title, character, comic, book, series, created, bob, rozakis, stephen, published, dc, comics, series, ran, twelve, issues, additional, speci...</td>\n",
       "      <td>[(mazing, man), (man, is), (is, the), (the, title), (title, character), (character, of), (of, a), (a, comic), (comic, book), (book, series), (seri...</td>\n",
       "      <td>[(title, character), (comic, book), (book, series), (series, created), (bob, rozakis), (dc, comics), (series, ran), (twelve, issues), (additional,...</td>\n",
       "      <td>[title character, comic book, book series, series created, bob rozakis, dc comics, series ran, twelve issues, additional special, special issues, ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>711 (Quality Comics)</td>\n",
       "      <td>is a fictional superhero from the golden age of comics he was created by george brenner and published by quality comics first appeared in police c...</td>\n",
       "      <td>[fictional, superhero, golden, age, comics, created, george, published, quality, comics, first, appeared, police, comics, august, lasted, january,...</td>\n",
       "      <td>[(is, a), (a, fictional), (fictional, superhero), (superhero, from), (from, the), (the, golden), (golden, age), (age, of), (of, comics), (comics, ...</td>\n",
       "      <td>[(fictional, superhero), (golden, age), (quality, comics), (comics, first), (first, appeared), (police, comics), (comics, august), (killed, fictio...</td>\n",
       "      <td>[fictional superhero, golden age, quality comics, comics first, first appeared, police comics, comics august, killed fictional, fictional characte...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Abigail Brand</td>\n",
       "      <td>special agent special agent abigail brand is a fictional character appearing in american comic book s published by marvel comics publication histo...</td>\n",
       "      <td>[special, agent, special, agent, abigail, brand, fictional, character, appearing, american, comic, book, published, marvel, comics, publication, h...</td>\n",
       "      <td>[(special, agent), (agent, special), (special, agent), (agent, abigail), (abigail, brand), (brand, is), (is, a), (a, fictional), (fictional, chara...</td>\n",
       "      <td>[(special, agent), (agent, special), (special, agent), (agent, abigail), (abigail, brand), (fictional, character), (character, appearing), (americ...</td>\n",
       "      <td>[special agent, agent special, special agent, agent abigail, abigail brand, fictional character, character appearing, american comic, comic book, ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Abin Sur</td>\n",
       "      <td>abin sur is a fictional character and a superhero from the dc comics dc universe he was a member of the green lantern corps and is best known as t...</td>\n",
       "      <td>[abin, sur, fictional, character, superhero, dc, comics, dc, universe, member, green, lantern, corps, best, known, predecessor, green, lantern, ha...</td>\n",
       "      <td>[(abin, sur), (sur, is), (is, a), (a, fictional), (fictional, character), (character, and), (and, a), (a, superhero), (superhero, from), (from, th...</td>\n",
       "      <td>[(abin, sur), (fictional, character), (dc, comics), (comics, dc), (dc, universe), (green, lantern), (lantern, corps), (best, known), (green, lante...</td>\n",
       "      <td>[abin sur, fictional character, dc comics, comics dc, dc universe, green lantern, lantern corps, best known, green lantern, lantern hal, hal jorda...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Abner Jenkins</td>\n",
       "      <td>abner ronald jenkins formerly known as the beetle comics beetle mach iv mach v mach vii and currently known as mach x and is a fictional character...</td>\n",
       "      <td>[abner, ronald, jenkins, formerly, known, beetle, comics, beetle, mach, mach, mach, vii, currently, known, mach, x, fictional, character, appearin...</td>\n",
       "      <td>[(abner, ronald), (ronald, jenkins), (jenkins, formerly), (formerly, known), (known, as), (as, the), (the, beetle), (beetle, comics), (comics, bee...</td>\n",
       "      <td>[(abner, ronald), (ronald, jenkins), (jenkins, formerly), (formerly, known), (beetle, comics), (comics, beetle), (beetle, mach), (mach, vii), (cur...</td>\n",
       "      <td>[abner ronald, ronald jenkins, jenkins formerly, formerly known, beetle comics, comics beetle, beetle mach, mach vii, currently known, mach x, fic...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   name  \\\n",
       "0           'Mazing Man   \n",
       "1  711 (Quality Comics)   \n",
       "2         Abigail Brand   \n",
       "3              Abin Sur   \n",
       "4         Abner Jenkins   \n",
       "\n",
       "                                                                                                                                             description  \\\n",
       "0  mazing man is the title character of a comic book series created by bob rozakis and stephen destefano and published by dc comics the series ran fo...   \n",
       "1  is a fictional superhero from the golden age of comics he was created by george brenner and published by quality comics first appeared in police c...   \n",
       "2  special agent special agent abigail brand is a fictional character appearing in american comic book s published by marvel comics publication histo...   \n",
       "3  abin sur is a fictional character and a superhero from the dc comics dc universe he was a member of the green lantern corps and is best known as t...   \n",
       "4  abner ronald jenkins formerly known as the beetle comics beetle mach iv mach v mach vii and currently known as mach x and is a fictional character...   \n",
       "\n",
       "                                                                                                                                              main_words  \\\n",
       "0  [man, title, character, comic, book, series, created, bob, rozakis, stephen, published, dc, comics, series, ran, twelve, issues, additional, speci...   \n",
       "1  [fictional, superhero, golden, age, comics, created, george, published, quality, comics, first, appeared, police, comics, august, lasted, january,...   \n",
       "2  [special, agent, special, agent, abigail, brand, fictional, character, appearing, american, comic, book, published, marvel, comics, publication, h...   \n",
       "3  [abin, sur, fictional, character, superhero, dc, comics, dc, universe, member, green, lantern, corps, best, known, predecessor, green, lantern, ha...   \n",
       "4  [abner, ronald, jenkins, formerly, known, beetle, comics, beetle, mach, mach, mach, vii, currently, known, mach, x, fictional, character, appearin...   \n",
       "\n",
       "                                                                                                                                                 bigrams  \\\n",
       "0  [(mazing, man), (man, is), (is, the), (the, title), (title, character), (character, of), (of, a), (a, comic), (comic, book), (book, series), (seri...   \n",
       "1  [(is, a), (a, fictional), (fictional, superhero), (superhero, from), (from, the), (the, golden), (golden, age), (age, of), (of, comics), (comics, ...   \n",
       "2  [(special, agent), (agent, special), (special, agent), (agent, abigail), (abigail, brand), (brand, is), (is, a), (a, fictional), (fictional, chara...   \n",
       "3  [(abin, sur), (sur, is), (is, a), (a, fictional), (fictional, character), (character, and), (and, a), (a, superhero), (superhero, from), (from, th...   \n",
       "4  [(abner, ronald), (ronald, jenkins), (jenkins, formerly), (formerly, known), (known, as), (as, the), (the, beetle), (beetle, comics), (comics, bee...   \n",
       "\n",
       "                                                                                                                                           clean_bigrams  \\\n",
       "0  [(title, character), (comic, book), (book, series), (series, created), (bob, rozakis), (dc, comics), (series, ran), (twelve, issues), (additional,...   \n",
       "1  [(fictional, superhero), (golden, age), (quality, comics), (comics, first), (first, appeared), (police, comics), (comics, august), (killed, fictio...   \n",
       "2  [(special, agent), (agent, special), (special, agent), (agent, abigail), (abigail, brand), (fictional, character), (character, appearing), (americ...   \n",
       "3  [(abin, sur), (fictional, character), (dc, comics), (comics, dc), (dc, universe), (green, lantern), (lantern, corps), (best, known), (green, lante...   \n",
       "4  [(abner, ronald), (ronald, jenkins), (jenkins, formerly), (formerly, known), (beetle, comics), (comics, beetle), (beetle, mach), (mach, vii), (cur...   \n",
       "\n",
       "                                                                                                                                        all_collocations  \n",
       "0  [title character, comic book, book series, series created, bob rozakis, dc comics, series ran, twelve issues, additional special, special issues, ...  \n",
       "1  [fictional superhero, golden age, quality comics, comics first, first appeared, police comics, comics august, killed fictional, fictional characte...  \n",
       "2  [special agent, agent special, special agent, agent abigail, abigail brand, fictional character, character appearing, american comic, comic book, ...  \n",
       "3  [abin sur, fictional character, dc comics, comics dc, dc universe, green lantern, lantern corps, best known, green lantern, lantern hal, hal jorda...  \n",
       "4  [abner ronald, ronald jenkins, jenkins formerly, formerly known, beetle comics, comics beetle, beetle mach, mach vii, currently known, mach x, fic...  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Cantidad de bigramas en el corpus:  1080272\n",
      "\n",
      "Bigramas más populares:\n",
      " [('x men', 9076), ('marvel comics', 8656), ('dc comics', 6009), ('spider man', 5192), ('category fictional', 3822), ('comic book', 3374), ('justice league', 3340), ('comics characters', 2944), ('fictional character', 2562), ('green lantern', 2419), ('category marvel', 2335), ('category comics', 2223), ('characters created', 2146), ('category characters', 2137), ('captain america', 2064), ('category dc', 1810), ('characters introduced', 1793), ('superheroes category', 1749), ('comics superheroes', 1649), ('tv series', 1566), ('character biography', 1537), ('iron man', 1522), ('first appeared', 1474), ('fantastic four', 1425), ('teen titans', 1400), ('uncanny x', 1387), ('external links', 1377), ('wonder woman', 1300), ('super heroes', 1160), ('new york', 1097), ('publication history', 1045), ('captain marvel', 1016), ('marvel universe', 1004), ('civil war', 963), ('new mutants', 908), ('superhuman strength', 903), ('x factor', 889), ('x force', 869), ('fictional characters', 856), ('dc universe', 847), ('lantern corps', 839), ('golden age', 824), ('playable character', 792), ('american comic', 784), ('justice society', 781), ('ghost rider', 780), ('united states', 762), ('amazing spider', 758), ('world war', 712), ('green arrow', 712)]\n"
     ]
    }
   ],
   "source": [
    "comicsDf[\"all_collocations\"] = list(map(lambda doc, bigrams: \n",
    "                            [b[0]+\" \"+b[1] for b in bigrams if b[0]+\" \"+b[1] in doc], \n",
    "                            comicsDf.description, comicsDf.clean_bigrams))\n",
    "display(comicsDf.head())\n",
    "\n",
    "clean_collocations = []\n",
    "for row in comicsDf.all_collocations:\n",
    "    clean_collocations.extend(row)\n",
    "common_clean_collocations = nltk.FreqDist(clean_collocations)\n",
    "\n",
    "print(\"Cantidad de bigramas en el corpus: \", common_clean_collocations.N())\n",
    "print(\"\\nBigramas más populares:\\n\", common_clean_collocations.most_common(50))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "425664 colocaciones\n",
      " [('x men', 9076), ('marvel comics', 8656), ('dc comics', 6009), ('spider man', 5192), ('category fictional', 3822), ('comic book', 3374), ('justice league', 3340), ('comics characters', 2944), ('fictional character', 2562), ('green lantern', 2419), ('category marvel', 2335), ('category comics', 2223), ('characters created', 2146), ('category characters', 2137), ('captain america', 2064), ('category dc', 1810), ('characters introduced', 1793), ('superheroes category', 1749), ('comics superheroes', 1649), ('tv series', 1566)] \n",
      "...\n",
      " [('exiles marvel', 83), ('alternate versions', 83), ('comics hercules', 83), ('marvel cinematic', 83), ('smallville season', 83), ('tiger shark', 83), ('diana prince', 83), ('idea mechanics', 82), ('writer artist', 82), ('history created', 82), ('comics exiles', 82), ('comics fear', 82), ('paperback comics', 82), ('life force', 82), ('mind controlled', 82), ('fun comics', 82), ('comics iceman', 82), ('film justice', 81), ('beyond gotham', 81), ('shapeshifters category', 81)] \n",
      "...\n",
      " [('continent zor', 1), ('new volcano', 1), ('rising core', 1), ('core pressure', 1), ('brainiac jor', 1), ('convince general', 1), ('zod commissioner', 1), ('commissioner zod', 1), ('rao beam', 1), ('vocal opponents', 1), ('dictator knowing', 1), ('zod plans', 1), ('krypton though', 1), ('though anderson', 1), ('reveal whether', 1), ('whether zor', 1), ('successful external', 1), ('links zor', 1), ('supermanica smallville', 1), ('el profile', 1)]\n"
     ]
    }
   ],
   "source": [
    "#colocaciones únicas\n",
    "nc = len(common_clean_collocations.most_common())\n",
    "print(nc, \"colocaciones\\n\", \n",
    "      common_clean_collocations.most_common(20),\n",
    "      \"\\n...\\n\",\n",
    "      list(common_clean_collocations.most_common())[int(nc/500):int(nc/500)+20],\n",
    "      \"\\n...\\n\",\n",
    "      list(common_clean_collocations.most_common())[nc-20:])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "A partir de este listado, con las 20 colocaciones más frecuentes, las 20 colocaciones menos frecuentes y 20 colocaciones con frecuencia media, observamos que no todas pueden tomarse como tokens, e incluso, la mayoría no tienen sentido como unidades. Construimos entonces las colocaciones tomando en cuenta el diccionario preliminar, reflejado en la columna <code>main_words</code> y mantenemos el nombre del personaje. Construimos una nueva columna con la descripción reescrita utilizando las colocaciones."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "import math\n",
    "\n",
    "comics_all_collocations = list(common_clean_collocations.keys())\n",
    "dict_all_collocations = dict(zip(comics_all_collocations, [0]*nc))\n",
    "for w in comics_all_collocations:\n",
    "    for d in comicsDf.description:\n",
    "        if(w in d):\n",
    "            dict_all_collocations[w] += 1\n",
    "\n",
    "top_collocations = []\n",
    "numDocs_comics = len(comicsDf)\n",
    "for d in comicsDf.description:\n",
    "    N = len(d.split()) - 1 #El número de bigramas es el número de palabras - 1\n",
    "    for w in reversed(comics_all_collocations): #En reversa para evitar problemas con los índices\n",
    "        if(w in d):\n",
    "            tfidf = d.count(w) / N * math.log(numDocs_comics/dict_all_collocations[w], 2)\n",
    "            if(tfidf > 0.01):\n",
    "                top_collocations.append(w)\n",
    "                comics_all_collocations.remove(w)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>name</th>\n",
       "      <th>description</th>\n",
       "      <th>main_words</th>\n",
       "      <th>bigrams</th>\n",
       "      <th>clean_bigrams</th>\n",
       "      <th>all_collocations</th>\n",
       "      <th>new_description</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>'Mazing Man</td>\n",
       "      <td>mazing man is the title character of a comic book series created by bob rozakis and stephen destefano and published by dc comics the series ran fo...</td>\n",
       "      <td>[man, title, character, comic, book, series, created, bob, rozakis, stephen, published, dc, comics, series, ran, twelve, issues, additional, speci...</td>\n",
       "      <td>[(mazing, man), (man, is), (is, the), (the, title), (title, character), (character, of), (of, a), (a, comic), (comic, book), (book, series), (seri...</td>\n",
       "      <td>[(title, character), (comic, book), (book, series), (series, created), (bob, rozakis), (dc, comics), (series, ran), (twelve, issues), (additional,...</td>\n",
       "      <td>[title character, comic book, book series, series created, bob rozakis, dc comics, series ran, twelve issues, additional special, special issues, ...</td>\n",
       "      <td>mazing_man man title_character comic_book series_created bob_rozakis stephen published dc_comics series ran twelve_issues additional_special_issue...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>711 (Quality Comics)</td>\n",
       "      <td>is a fictional superhero from the golden age of comics he was created by george brenner and published by quality comics first appeared in police c...</td>\n",
       "      <td>[fictional, superhero, golden, age, comics, created, george, published, quality, comics, first, appeared, police, comics, august, lasted, january,...</td>\n",
       "      <td>[(is, a), (a, fictional), (fictional, superhero), (superhero, from), (from, the), (the, golden), (golden, age), (age, of), (of, comics), (comics, ...</td>\n",
       "      <td>[(fictional, superhero), (golden, age), (quality, comics), (comics, first), (first, appeared), (police, comics), (comics, august), (killed, fictio...</td>\n",
       "      <td>[fictional superhero, golden age, quality comics, comics first, first appeared, police comics, comics august, killed fictional, fictional characte...</td>\n",
       "      <td>711_quality_comics fictional superhero golden_age comics_created george_published quality_comics first_appeared police_comics_august lasted januar...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Abigail Brand</td>\n",
       "      <td>special agent special agent abigail brand is a fictional character appearing in american comic book s published by marvel comics publication histo...</td>\n",
       "      <td>[special, agent, special, agent, abigail, brand, fictional, character, appearing, american, comic, book, published, marvel, comics, publication, h...</td>\n",
       "      <td>[(special, agent), (agent, special), (special, agent), (agent, abigail), (abigail, brand), (brand, is), (is, a), (a, fictional), (fictional, chara...</td>\n",
       "      <td>[(special, agent), (agent, special), (special, agent), (agent, abigail), (abigail, brand), (fictional, character), (character, appearing), (americ...</td>\n",
       "      <td>[special agent, agent special, special agent, agent abigail, abigail brand, fictional character, character appearing, american comic, comic book, ...</td>\n",
       "      <td>abigail_brand special agent special agent abigail_brand fictional character_appearing american_comic_book published marvel_comics publication hist...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Abin Sur</td>\n",
       "      <td>abin sur is a fictional character and a superhero from the dc comics dc universe he was a member of the green lantern corps and is best known as t...</td>\n",
       "      <td>[abin, sur, fictional, character, superhero, dc, comics, dc, universe, member, green, lantern, corps, best, known, predecessor, green, lantern, ha...</td>\n",
       "      <td>[(abin, sur), (sur, is), (is, a), (a, fictional), (fictional, character), (character, and), (and, a), (a, superhero), (superhero, from), (from, th...</td>\n",
       "      <td>[(abin, sur), (fictional, character), (dc, comics), (comics, dc), (dc, universe), (green, lantern), (lantern, corps), (best, known), (green, lante...</td>\n",
       "      <td>[abin sur, fictional character, dc comics, comics dc, dc universe, green lantern, lantern corps, best known, green lantern, lantern hal, hal jorda...</td>\n",
       "      <td>abin_sur abin sur fictional character_superhero dc_comics dc_universe member green_lantern corps best_known predecessor green_lantern hal_jordan a...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Abner Jenkins</td>\n",
       "      <td>abner ronald jenkins formerly known as the beetle comics beetle mach iv mach v mach vii and currently known as mach x and is a fictional character...</td>\n",
       "      <td>[abner, ronald, jenkins, formerly, known, beetle, comics, beetle, mach, mach, mach, vii, currently, known, mach, x, fictional, character, appearin...</td>\n",
       "      <td>[(abner, ronald), (ronald, jenkins), (jenkins, formerly), (formerly, known), (known, as), (as, the), (the, beetle), (beetle, comics), (comics, bee...</td>\n",
       "      <td>[(abner, ronald), (ronald, jenkins), (jenkins, formerly), (formerly, known), (beetle, comics), (comics, beetle), (beetle, mach), (mach, vii), (cur...</td>\n",
       "      <td>[abner ronald, ronald jenkins, jenkins formerly, formerly known, beetle comics, comics beetle, beetle mach, mach vii, currently known, mach x, fic...</td>\n",
       "      <td>abner_jenkins abner ronald jenkins formerly_known beetle_comics beetle mach mach mach_vii currently known mach x_fictional character_appearing ame...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   name  \\\n",
       "0           'Mazing Man   \n",
       "1  711 (Quality Comics)   \n",
       "2         Abigail Brand   \n",
       "3              Abin Sur   \n",
       "4         Abner Jenkins   \n",
       "\n",
       "                                                                                                                                             description  \\\n",
       "0  mazing man is the title character of a comic book series created by bob rozakis and stephen destefano and published by dc comics the series ran fo...   \n",
       "1  is a fictional superhero from the golden age of comics he was created by george brenner and published by quality comics first appeared in police c...   \n",
       "2  special agent special agent abigail brand is a fictional character appearing in american comic book s published by marvel comics publication histo...   \n",
       "3  abin sur is a fictional character and a superhero from the dc comics dc universe he was a member of the green lantern corps and is best known as t...   \n",
       "4  abner ronald jenkins formerly known as the beetle comics beetle mach iv mach v mach vii and currently known as mach x and is a fictional character...   \n",
       "\n",
       "                                                                                                                                              main_words  \\\n",
       "0  [man, title, character, comic, book, series, created, bob, rozakis, stephen, published, dc, comics, series, ran, twelve, issues, additional, speci...   \n",
       "1  [fictional, superhero, golden, age, comics, created, george, published, quality, comics, first, appeared, police, comics, august, lasted, january,...   \n",
       "2  [special, agent, special, agent, abigail, brand, fictional, character, appearing, american, comic, book, published, marvel, comics, publication, h...   \n",
       "3  [abin, sur, fictional, character, superhero, dc, comics, dc, universe, member, green, lantern, corps, best, known, predecessor, green, lantern, ha...   \n",
       "4  [abner, ronald, jenkins, formerly, known, beetle, comics, beetle, mach, mach, mach, vii, currently, known, mach, x, fictional, character, appearin...   \n",
       "\n",
       "                                                                                                                                                 bigrams  \\\n",
       "0  [(mazing, man), (man, is), (is, the), (the, title), (title, character), (character, of), (of, a), (a, comic), (comic, book), (book, series), (seri...   \n",
       "1  [(is, a), (a, fictional), (fictional, superhero), (superhero, from), (from, the), (the, golden), (golden, age), (age, of), (of, comics), (comics, ...   \n",
       "2  [(special, agent), (agent, special), (special, agent), (agent, abigail), (abigail, brand), (brand, is), (is, a), (a, fictional), (fictional, chara...   \n",
       "3  [(abin, sur), (sur, is), (is, a), (a, fictional), (fictional, character), (character, and), (and, a), (a, superhero), (superhero, from), (from, th...   \n",
       "4  [(abner, ronald), (ronald, jenkins), (jenkins, formerly), (formerly, known), (known, as), (as, the), (the, beetle), (beetle, comics), (comics, bee...   \n",
       "\n",
       "                                                                                                                                           clean_bigrams  \\\n",
       "0  [(title, character), (comic, book), (book, series), (series, created), (bob, rozakis), (dc, comics), (series, ran), (twelve, issues), (additional,...   \n",
       "1  [(fictional, superhero), (golden, age), (quality, comics), (comics, first), (first, appeared), (police, comics), (comics, august), (killed, fictio...   \n",
       "2  [(special, agent), (agent, special), (special, agent), (agent, abigail), (abigail, brand), (fictional, character), (character, appearing), (americ...   \n",
       "3  [(abin, sur), (fictional, character), (dc, comics), (comics, dc), (dc, universe), (green, lantern), (lantern, corps), (best, known), (green, lante...   \n",
       "4  [(abner, ronald), (ronald, jenkins), (jenkins, formerly), (formerly, known), (beetle, comics), (comics, beetle), (beetle, mach), (mach, vii), (cur...   \n",
       "\n",
       "                                                                                                                                        all_collocations  \\\n",
       "0  [title character, comic book, book series, series created, bob rozakis, dc comics, series ran, twelve issues, additional special, special issues, ...   \n",
       "1  [fictional superhero, golden age, quality comics, comics first, first appeared, police comics, comics august, killed fictional, fictional characte...   \n",
       "2  [special agent, agent special, special agent, agent abigail, abigail brand, fictional character, character appearing, american comic, comic book, ...   \n",
       "3  [abin sur, fictional character, dc comics, comics dc, dc universe, green lantern, lantern corps, best known, green lantern, lantern hal, hal jorda...   \n",
       "4  [abner ronald, ronald jenkins, jenkins formerly, formerly known, beetle comics, comics beetle, beetle mach, mach vii, currently known, mach x, fic...   \n",
       "\n",
       "                                                                                                                                         new_description  \n",
       "0  mazing_man man title_character comic_book series_created bob_rozakis stephen published dc_comics series ran twelve_issues additional_special_issue...  \n",
       "1  711_quality_comics fictional superhero golden_age comics_created george_published quality_comics first_appeared police_comics_august lasted januar...  \n",
       "2  abigail_brand special agent special agent abigail_brand fictional character_appearing american_comic_book published marvel_comics publication hist...  \n",
       "3  abin_sur abin sur fictional character_superhero dc_comics dc_universe member green_lantern corps best_known predecessor green_lantern hal_jordan a...  \n",
       "4  abner_jenkins abner ronald jenkins formerly_known beetle_comics beetle mach mach mach_vii currently known mach x_fictional character_appearing ame...  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "collocations_tokens = [b for b in common_clean_collocations.most_common() \n",
    "                      if b[0] in top_collocations]\n",
    "comicsDf = comicsDf.reindex(columns = [\"name\", \"description\", \"main_words\", \"bigrams\", \n",
    "                                        \"clean_bigrams\", \"all_collocations\", \"new_description\"])\n",
    "\n",
    "for i, row in zip(range(len(comicsDf)), comicsDf.main_words):\n",
    "    #Eliminar símbolos no alfanuméricos en el nombre\n",
    "    n = re.sub(\"\\.\", \"\", comicsDf.loc[i, \"name\"])\n",
    "    n = re.sub(r\"-\", \"_\", n)\n",
    "    n = \"_\".join(re.sub(\"[\\W]\", \" \", n).split()) + \" \"\n",
    "\n",
    "    #reemplazar colocaciones por tokens unificados y eliminar palabras poco utilizadas\n",
    "    s = \" \".join(row)\n",
    "    for w in collocations_tokens:\n",
    "        s = re.sub(\" \" + w[0], \" \" + \"_\".join(w[0].split()), s)\n",
    "    s = \" \".join(s.split())\n",
    "\n",
    "    #Agregar el título de la página en forma de token para rescatar el tema central \n",
    "    #que puede perderse como palabra sin utilidad\n",
    "    comicsDf.loc[i, \"new_description\"] = n.lower() + s\n",
    "\n",
    "display(comicsDf.head())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "comicsDf.to_json(\"Data Sets/Comics/lexicon_comics.json\", orient='records') "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "A continuación, analizamos nuevamente el universo de palabras usadas en el corpus:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Cantidad de tokens en el corpus:  90132\n",
      "\n",
      "Tokens más populares:\n",
      "['x_men', 'character', 'comics', 'one', 'time', 'earth', 'marvel_comics', 'also', 'batman', 'spider_man', 'series', 'powers', 'appears', 'team', 'later', 'justice_league', 'however', 'superman', 'member', 'battle', 'dc_comics', 'revealed', 'new', 'death', 'avengers', 'created', 'father', 'version', 'introduced', 'characters', 'body', 'help', 'able', 'would', 'power', 'vol', 'killed', 'green_lantern', 'life', 'back', 'part', 'two', 'fight', 'voiced', 'first', 'fictional', 'wolverine', 'use', 'ability', 'used']\n",
      "...\n",
      "['steve_trevor', 'licensed', 'lied', 'user', 'pat', 'kole', 'trench', 'new_jersey', 'secret_warriors', 'salaak', 'vigilantes', 'boston', 'flash_vol', 'aphrodite', 'lightning_saga', 'neutralize', 'miniature', 'jumps', 'dormant', 'sunfire_comics', 'epic', 'anti_life_equation', 'captain_cold', 'carrie', 'selina_kyle', 'youngest', 'depressed', 'gray', 'comic_strip', 'lori', 'captain_dynamo', 'kents', 'pepper_potts', 'doctor_occult', 'moon_girl', 'squadron_sinister', 'makkari', 'witchblade', 'forbidden', 'break_free', 'would_take', 'interstellar', 'unsure', 'meteor', 'completing', 'bite', 'distinct', 'andy_lanning', 'breach', 'archenemy']\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA5MAAAGjCAYAAACxCkENAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nSUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x15cd947f0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA6UAAAGjCAYAAADU57NyAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\ngsr4/9hr+NfgY1o3AAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x165284a90>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "new_main_words = []\n",
    "for row in comicsDf.new_description:\n",
    "    new_main_words.extend(row.split())\n",
    "common_new_main_words = nltk.FreqDist(new_main_words)\n",
    "\n",
    "print(\"Cantidad de tokens en el corpus: \", len(common_new_main_words.most_common()))\n",
    "\n",
    "lexicon_comics = [w[0] for w in common_new_main_words.most_common(6000)]\n",
    "print(\"\\nTokens más populares:\\n{}\\n...\\n{}\"\n",
    "      .format(lexicon_comics[:50], lexicon_comics[5950:6000]))\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "plt.rcParams.update({'font.size': 14})\n",
    "plt.figure(figsize=(15, 5))\n",
    "common_new_main_words.plot(50, cumulative=False)\n",
    "plt.figure(figsize=(15, 5))\n",
    "common_new_main_words.plot(50, cumulative=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "De aquí se observa una simplificación del lenguaje. A partir de este punto, podemos volver a realizar el análisis para reducir el tamaño del conjunto de tokens utilizados. \n",
    "\n",
    "A partir de un procedimiento como el señalado, es posible automatizar la creación de un lexicón, aunque los últimos pasos tienen que hacerse con la intervención de un experto. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Conclusiones\n",
    "\n",
    "El uso de $N$-Gramas y *colocaciones* permite anticipar el uso de palabras y aumentar la precisión del análisis léxico. Es de destacarse que esta aproximación agrega a la etapa del análisis léxico, elementos del análisis sintáctico e incluso del análisis semántico, al establecer un contexto a la ocurrencia de las palabras e incluso darle sentido a las palabras cuando vienen acompañadas por otras palabras (que definen el contexto).\n",
    "\n",
    "<hr style=\"border-width: 3px;\">\n",
    "\n",
    "### Tarea 4\n",
    "\n",
    "Realice las operaciones presentadas en esta lección para mejorar la construcción de su lexicón.\n",
    "\n",
    "**Fecha de entrega**: Martes 5 de septiembre."
   ]
  }
 ],
 "metadata": {
  "celltoolbar": "Raw Cell Format",
  "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": "3.6.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}