3. N - Gram Mode(Tri Gram & Word Gram).ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# N - GRAM Model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "1.1 Character N-Gram Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import random"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Sample data\n",
    "text = \"\"\"Global warming or climate change has become a worldwide concern. It is gradually developing into an unprecedented environmental crisis evident in melting glaciers, changing weather patterns, rising sea levels, floods, cyclones and droughts. Global warming implies an increase in the average temperature of the Earth due to entrapment of greenhouse gases in the earth’s atmosphere.\"\"\"\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Order of the grams\n",
    "n = 6"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "ngrams = {}\n",
    "\n",
    "#create the n-grams\n",
    "for i in range(len(text) - n):\n",
    "    gram = text[i:i+n] # text[0;3] = Glo\n",
    "    if gram not in ngrams.keys():\n",
    "        ngrams[gram] = []\n",
    "    ngrams[gram].append(text[i+n])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'Global': [' ', ' '],\n",
       " 'lobal ': ['w', 'w'],\n",
       " 'obal w': ['a', 'a'],\n",
       " 'bal wa': ['r', 'r'],\n",
       " 'al war': ['m', 'm'],\n",
       " 'l warm': ['i', 'i'],\n",
       " ' warmi': ['n', 'n'],\n",
       " 'warmin': ['g', 'g'],\n",
       " 'arming': [' ', ' '],\n",
       " 'rming ': ['o', 'i'],\n",
       " 'ming o': ['r'],\n",
       " 'ing or': [' '],\n",
       " 'ng or ': ['c'],\n",
       " 'g or c': ['l'],\n",
       " ' or cl': ['i'],\n",
       " 'or cli': ['m'],\n",
       " 'r clim': ['a'],\n",
       " ' clima': ['t'],\n",
       " 'climat': ['e'],\n",
       " 'limate': [' '],\n",
       " 'imate ': ['c'],\n",
       " 'mate c': ['h'],\n",
       " 'ate ch': ['a'],\n",
       " 'te cha': ['n'],\n",
       " 'e chan': ['g'],\n",
       " ' chang': ['e', 'i'],\n",
       " 'change': [' '],\n",
       " 'hange ': ['h'],\n",
       " 'ange h': ['a'],\n",
       " 'nge ha': ['s'],\n",
       " 'ge has': [' '],\n",
       " 'e has ': ['b'],\n",
       " ' has b': ['e'],\n",
       " 'has be': ['c'],\n",
       " 'as bec': ['o'],\n",
       " 's beco': ['m'],\n",
       " ' becom': ['e'],\n",
       " 'become': [' '],\n",
       " 'ecome ': ['a'],\n",
       " 'come a': [' '],\n",
       " 'ome a ': ['w'],\n",
       " 'me a w': ['o'],\n",
       " 'e a wo': ['r'],\n",
       " ' a wor': ['l'],\n",
       " 'a worl': ['d'],\n",
       " ' world': ['w'],\n",
       " 'worldw': ['i'],\n",
       " 'orldwi': ['d'],\n",
       " 'rldwid': ['e'],\n",
       " 'ldwide': [' '],\n",
       " 'dwide ': ['c'],\n",
       " 'wide c': ['o'],\n",
       " 'ide co': ['n'],\n",
       " 'de con': ['c'],\n",
       " 'e conc': ['e'],\n",
       " ' conce': ['r'],\n",
       " 'concer': ['n'],\n",
       " 'oncern': ['.'],\n",
       " 'ncern.': [' '],\n",
       " 'cern. ': ['I'],\n",
       " 'ern. I': ['t'],\n",
       " 'rn. It': [' '],\n",
       " 'n. It ': ['i'],\n",
       " '. It i': ['s'],\n",
       " ' It is': [' '],\n",
       " 'It is ': ['g'],\n",
       " 't is g': ['r'],\n",
       " ' is gr': ['a'],\n",
       " 'is gra': ['d'],\n",
       " 's grad': ['u'],\n",
       " ' gradu': ['a'],\n",
       " 'gradua': ['l'],\n",
       " 'radual': ['l'],\n",
       " 'aduall': ['y'],\n",
       " 'dually': [' '],\n",
       " 'ually ': ['d'],\n",
       " 'ally d': ['e'],\n",
       " 'lly de': ['v'],\n",
       " 'ly dev': ['e'],\n",
       " 'y deve': ['l'],\n",
       " ' devel': ['o'],\n",
       " 'develo': ['p'],\n",
       " 'evelop': ['i'],\n",
       " 'velopi': ['n'],\n",
       " 'elopin': ['g'],\n",
       " 'loping': [' '],\n",
       " 'oping ': ['i'],\n",
       " 'ping i': ['n'],\n",
       " 'ing in': ['t'],\n",
       " 'ng int': ['o'],\n",
       " 'g into': [' '],\n",
       " ' into ': ['a'],\n",
       " 'into a': ['n'],\n",
       " 'nto an': [' '],\n",
       " 'to an ': ['u'],\n",
       " 'o an u': ['n'],\n",
       " ' an un': ['p'],\n",
       " 'an unp': ['r'],\n",
       " 'n unpr': ['e'],\n",
       " ' unpre': ['c'],\n",
       " 'unprec': ['e'],\n",
       " 'nprece': ['d'],\n",
       " 'preced': ['e'],\n",
       " 'recede': ['n'],\n",
       " 'eceden': ['t'],\n",
       " 'cedent': ['e'],\n",
       " 'edente': ['d'],\n",
       " 'dented': [' '],\n",
       " 'ented ': ['e'],\n",
       " 'nted e': ['n'],\n",
       " 'ted en': ['v'],\n",
       " 'ed env': ['i'],\n",
       " 'd envi': ['r'],\n",
       " ' envir': ['o'],\n",
       " 'enviro': ['n'],\n",
       " 'nviron': ['m'],\n",
       " 'vironm': ['e'],\n",
       " 'ironme': ['n'],\n",
       " 'ronmen': ['t'],\n",
       " 'onment': ['a'],\n",
       " 'nmenta': ['l'],\n",
       " 'mental': [' '],\n",
       " 'ental ': ['c'],\n",
       " 'ntal c': ['r'],\n",
       " 'tal cr': ['i'],\n",
       " 'al cri': ['s'],\n",
       " 'l cris': ['i'],\n",
       " ' crisi': ['s'],\n",
       " 'crisis': [' '],\n",
       " 'risis ': ['e'],\n",
       " 'isis e': ['v'],\n",
       " 'sis ev': ['i'],\n",
       " 'is evi': ['d'],\n",
       " 's evid': ['e'],\n",
       " ' evide': ['n'],\n",
       " 'eviden': ['t'],\n",
       " 'vident': [' '],\n",
       " 'ident ': ['i'],\n",
       " 'dent i': ['n'],\n",
       " 'ent in': [' '],\n",
       " 'nt in ': ['m'],\n",
       " 't in m': ['e'],\n",
       " ' in me': ['l'],\n",
       " 'in mel': ['t'],\n",
       " 'n melt': ['i'],\n",
       " ' melti': ['n'],\n",
       " 'meltin': ['g'],\n",
       " 'elting': [' '],\n",
       " 'lting ': ['g'],\n",
       " 'ting g': ['l'],\n",
       " 'ing gl': ['a'],\n",
       " 'ng gla': ['c'],\n",
       " 'g glac': ['i'],\n",
       " ' glaci': ['e'],\n",
       " 'glacie': ['r'],\n",
       " 'lacier': ['s'],\n",
       " 'aciers': [','],\n",
       " 'ciers,': [' '],\n",
       " 'iers, ': ['c'],\n",
       " 'ers, c': ['h'],\n",
       " 'rs, ch': ['a'],\n",
       " 's, cha': ['n'],\n",
       " ', chan': ['g'],\n",
       " 'changi': ['n'],\n",
       " 'hangin': ['g'],\n",
       " 'anging': [' '],\n",
       " 'nging ': ['w'],\n",
       " 'ging w': ['e'],\n",
       " 'ing we': ['a'],\n",
       " 'ng wea': ['t'],\n",
       " 'g weat': ['h'],\n",
       " ' weath': ['e'],\n",
       " 'weathe': ['r'],\n",
       " 'eather': [' '],\n",
       " 'ather ': ['p'],\n",
       " 'ther p': ['a'],\n",
       " 'her pa': ['t'],\n",
       " 'er pat': ['t'],\n",
       " 'r patt': ['e'],\n",
       " ' patte': ['r'],\n",
       " 'patter': ['n'],\n",
       " 'attern': ['s'],\n",
       " 'tterns': [','],\n",
       " 'terns,': [' '],\n",
       " 'erns, ': ['r'],\n",
       " 'rns, r': ['i'],\n",
       " 'ns, ri': ['s'],\n",
       " 's, ris': ['i'],\n",
       " ', risi': ['n'],\n",
       " ' risin': ['g'],\n",
       " 'rising': [' '],\n",
       " 'ising ': ['s'],\n",
       " 'sing s': ['e'],\n",
       " 'ing se': ['a'],\n",
       " 'ng sea': [' '],\n",
       " 'g sea ': ['l'],\n",
       " ' sea l': ['e'],\n",
       " 'sea le': ['v'],\n",
       " 'ea lev': ['e'],\n",
       " 'a leve': ['l'],\n",
       " ' level': ['s'],\n",
       " 'levels': [','],\n",
       " 'evels,': [' '],\n",
       " 'vels, ': ['f'],\n",
       " 'els, f': ['l'],\n",
       " 'ls, fl': ['o'],\n",
       " 's, flo': ['o'],\n",
       " ', floo': ['d'],\n",
       " ' flood': ['s'],\n",
       " 'floods': [','],\n",
       " 'loods,': [' '],\n",
       " 'oods, ': ['c'],\n",
       " 'ods, c': ['y'],\n",
       " 'ds, cy': ['c'],\n",
       " 's, cyc': ['l'],\n",
       " ', cycl': ['o'],\n",
       " ' cyclo': ['n'],\n",
       " 'cyclon': ['e'],\n",
       " 'yclone': ['s'],\n",
       " 'clones': [' '],\n",
       " 'lones ': ['a'],\n",
       " 'ones a': ['n'],\n",
       " 'nes an': ['d'],\n",
       " 'es and': [' '],\n",
       " 's and ': ['d'],\n",
       " ' and d': ['r'],\n",
       " 'and dr': ['o'],\n",
       " 'nd dro': ['u'],\n",
       " 'd drou': ['g'],\n",
       " ' droug': ['h'],\n",
       " 'drough': ['t'],\n",
       " 'rought': ['s'],\n",
       " 'oughts': ['.'],\n",
       " 'ughts.': [' '],\n",
       " 'ghts. ': ['G'],\n",
       " 'hts. G': ['l'],\n",
       " 'ts. Gl': ['o'],\n",
       " 's. Glo': ['b'],\n",
       " '. Glob': ['a'],\n",
       " ' Globa': ['l'],\n",
       " 'ming i': ['m'],\n",
       " 'ing im': ['p'],\n",
       " 'ng imp': ['l'],\n",
       " 'g impl': ['i'],\n",
       " ' impli': ['e'],\n",
       " 'implie': ['s'],\n",
       " 'mplies': [' '],\n",
       " 'plies ': ['a'],\n",
       " 'lies a': ['n'],\n",
       " 'ies an': [' '],\n",
       " 'es an ': ['i'],\n",
       " 's an i': ['n'],\n",
       " ' an in': ['c'],\n",
       " 'an inc': ['r'],\n",
       " 'n incr': ['e'],\n",
       " ' incre': ['a'],\n",
       " 'increa': ['s'],\n",
       " 'ncreas': ['e'],\n",
       " 'crease': [' '],\n",
       " 'rease ': ['i'],\n",
       " 'ease i': ['n'],\n",
       " 'ase in': [' '],\n",
       " 'se in ': ['t'],\n",
       " 'e in t': ['h'],\n",
       " ' in th': ['e', 'e'],\n",
       " 'in the': [' ', ' '],\n",
       " 'n the ': ['a', 'e'],\n",
       " ' the a': ['v'],\n",
       " 'the av': ['e'],\n",
       " 'he ave': ['r'],\n",
       " 'e aver': ['a'],\n",
       " ' avera': ['g'],\n",
       " 'averag': ['e'],\n",
       " 'verage': [' '],\n",
       " 'erage ': ['t'],\n",
       " 'rage t': ['e'],\n",
       " 'age te': ['m'],\n",
       " 'ge tem': ['p'],\n",
       " 'e temp': ['e'],\n",
       " ' tempe': ['r'],\n",
       " 'temper': ['a'],\n",
       " 'empera': ['t'],\n",
       " 'mperat': ['u'],\n",
       " 'peratu': ['r'],\n",
       " 'eratur': ['e'],\n",
       " 'rature': [' '],\n",
       " 'ature ': ['o'],\n",
       " 'ture o': ['f'],\n",
       " 'ure of': [' '],\n",
       " 're of ': ['t'],\n",
       " 'e of t': ['h'],\n",
       " ' of th': ['e'],\n",
       " 'of the': [' '],\n",
       " 'f the ': ['E'],\n",
       " ' the E': ['a'],\n",
       " 'the Ea': ['r'],\n",
       " 'he Ear': ['t'],\n",
       " 'e Eart': ['h'],\n",
       " ' Earth': [' '],\n",
       " 'Earth ': ['d'],\n",
       " 'arth d': ['u'],\n",
       " 'rth du': ['e'],\n",
       " 'th due': [' '],\n",
       " 'h due ': ['t'],\n",
       " ' due t': ['o'],\n",
       " 'due to': [' '],\n",
       " 'ue to ': ['e'],\n",
       " 'e to e': ['n'],\n",
       " ' to en': ['t'],\n",
       " 'to ent': ['r'],\n",
       " 'o entr': ['a'],\n",
       " ' entra': ['p'],\n",
       " 'entrap': ['m'],\n",
       " 'ntrapm': ['e'],\n",
       " 'trapme': ['n'],\n",
       " 'rapmen': ['t'],\n",
       " 'apment': [' '],\n",
       " 'pment ': ['o'],\n",
       " 'ment o': ['f'],\n",
       " 'ent of': [' '],\n",
       " 'nt of ': ['g'],\n",
       " 't of g': ['r'],\n",
       " ' of gr': ['e'],\n",
       " 'of gre': ['e'],\n",
       " 'f gree': ['n'],\n",
       " ' green': ['h'],\n",
       " 'greenh': ['o'],\n",
       " 'reenho': ['u'],\n",
       " 'eenhou': ['s'],\n",
       " 'enhous': ['e'],\n",
       " 'nhouse': [' '],\n",
       " 'house ': ['g'],\n",
       " 'ouse g': ['a'],\n",
       " 'use ga': ['s'],\n",
       " 'se gas': ['e'],\n",
       " 'e gase': ['s'],\n",
       " ' gases': [' '],\n",
       " 'gases ': ['i'],\n",
       " 'ases i': ['n'],\n",
       " 'ses in': [' '],\n",
       " 'es in ': ['t'],\n",
       " 's in t': ['h'],\n",
       " ' the e': ['a'],\n",
       " 'the ea': ['r'],\n",
       " 'he ear': ['t'],\n",
       " 'e eart': ['h'],\n",
       " ' earth': ['’'],\n",
       " 'earth’': ['s'],\n",
       " 'arth’s': [' '],\n",
       " 'rth’s ': ['a'],\n",
       " 'th’s a': ['t'],\n",
       " 'h’s at': ['m'],\n",
       " '’s atm': ['o'],\n",
       " 's atmo': ['s'],\n",
       " ' atmos': ['p'],\n",
       " 'atmosp': ['h'],\n",
       " 'tmosph': ['e'],\n",
       " 'mosphe': ['r'],\n",
       " 'ospher': ['e'],\n",
       " 'sphere': ['.']}"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ngrams"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "360"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(ngrams)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Testing our N-Gram Model\n",
    "\n",
    "currentGram = text[0:n]\n",
    "result =currentGram\n",
    "for i in range(100):\n",
    "    if currentGram not in ngrams.keys():\n",
    "        break\n",
    "    possibility = ngrams[currentGram]\n",
    "    nextItem = possibility[random.randrange(len(possibility))]\n",
    "    result += nextItem\n",
    "    currentGram = result[len(result)-n:len(result)]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Global warming implies an increase in the average temperature of the Earth due to entrapment of greenhouse\n"
     ]
    }
   ],
   "source": [
    "print(result)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "1.2 Word N-Gram Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "import random\n",
    "import nltk"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Sample data\n",
    "text = \"\"\"Global warming or climate change has become a worldwide concern. It is gradually developing into an unprecedented environmental crisis evident in melting glaciers, changing weather patterns, rising sea levels, floods, cyclones and droughts. Global warming implies an increase in the average temperature of the Earth due to entrapment of greenhouse gases in the earth’s atmosphere.\"\"\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "n = 2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "ngrams = {}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Building the model\n",
    "words = nltk.word_tokenize(text)\n",
    "for i in range(len(words)-n):\n",
    "    gram = ' '.join(words[i:i+n])\n",
    "    if gram not in ngrams.keys():\n",
    "        ngrams[gram] = []\n",
    "    ngrams[gram].append(words[i+n])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'Global warming': ['or', 'implies'],\n",
       " 'warming or': ['climate'],\n",
       " 'or climate': ['change'],\n",
       " 'climate change': ['has'],\n",
       " 'change has': ['become'],\n",
       " 'has become': ['a'],\n",
       " 'become a': ['worldwide'],\n",
       " 'a worldwide': ['concern'],\n",
       " 'worldwide concern': ['.'],\n",
       " 'concern .': ['It'],\n",
       " '. It': ['is'],\n",
       " 'It is': ['gradually'],\n",
       " 'is gradually': ['developing'],\n",
       " 'gradually developing': ['into'],\n",
       " 'developing into': ['an'],\n",
       " 'into an': ['unprecedented'],\n",
       " 'an unprecedented': ['environmental'],\n",
       " 'unprecedented environmental': ['crisis'],\n",
       " 'environmental crisis': ['evident'],\n",
       " 'crisis evident': ['in'],\n",
       " 'evident in': ['melting'],\n",
       " 'in melting': ['glaciers'],\n",
       " 'melting glaciers': [','],\n",
       " 'glaciers ,': ['changing'],\n",
       " ', changing': ['weather'],\n",
       " 'changing weather': ['patterns'],\n",
       " 'weather patterns': [','],\n",
       " 'patterns ,': ['rising'],\n",
       " ', rising': ['sea'],\n",
       " 'rising sea': ['levels'],\n",
       " 'sea levels': [','],\n",
       " 'levels ,': ['floods'],\n",
       " ', floods': [','],\n",
       " 'floods ,': ['cyclones'],\n",
       " ', cyclones': ['and'],\n",
       " 'cyclones and': ['droughts'],\n",
       " 'and droughts': ['.'],\n",
       " 'droughts .': ['Global'],\n",
       " '. Global': ['warming'],\n",
       " 'warming implies': ['an'],\n",
       " 'implies an': ['increase'],\n",
       " 'an increase': ['in'],\n",
       " 'increase in': ['the'],\n",
       " 'in the': ['average', 'earth'],\n",
       " 'the average': ['temperature'],\n",
       " 'average temperature': ['of'],\n",
       " 'temperature of': ['the'],\n",
       " 'of the': ['Earth'],\n",
       " 'the Earth': ['due'],\n",
       " 'Earth due': ['to'],\n",
       " 'due to': ['entrapment'],\n",
       " 'to entrapment': ['of'],\n",
       " 'entrapment of': ['greenhouse'],\n",
       " 'of greenhouse': ['gases'],\n",
       " 'greenhouse gases': ['in'],\n",
       " 'gases in': ['the'],\n",
       " 'the earth': ['’'],\n",
       " 'earth ’': ['s'],\n",
       " '’ s': ['atmosphere'],\n",
       " 's atmosphere': ['.']}"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ngrams"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Global warming implies an increase in the earth ’ s atmosphere .\n"
     ]
    }
   ],
   "source": [
    "# Testing the model\n",
    "currentGram = ' '.join(words[0:n])\n",
    "result = currentGram\n",
    "for i in range(30):\n",
    "    if currentGram not in ngrams.keys():\n",
    "        break\n",
    "    possibilities = ngrams[currentGram]\n",
    "    nextItem = possibilities[random.randrange(len(possibilities))]\n",
    "    result += ' '+nextItem\n",
    "    rWords = nltk.word_tokenize(result)\n",
    "    currentGram = ' '.join(rWords[len(rWords)-n:len(rWords)])\n",
    "\n",
    "print(result)"
   ]
  },
  {
   "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": "3.6.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}