Benchmark testing of the gensim topic coherence pipeline on 20NG dataset

20ng_benchmark_testing.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "import re\n",
    "\n",
    "from sklearn.datasets import fetch_20newsgroups\n",
    "from scipy.stats import pearsonr\n",
    "from datetime import datetime\n",
    "\n",
    "from gensim.models import CoherenceModel\n",
    "from gensim.corpora.dictionary import Dictionary"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "dataset = fetch_20newsgroups(subset='all', remove=('headers', 'footers', 'quotes'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "documents = dataset['data']  # is a list of documents"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "texts = []\n",
    "for document in documents:\n",
    "    # lower case all words\n",
    "    lowered = document.lower()\n",
    "    # remove punctuation and split into seperate words\n",
    "    words = re.findall(r'\\w+', lowered, flags = re.UNICODE | re.LOCALE)\n",
    "    texts.append(words)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "dictionary = Dictionary(texts)\n",
    "corpus = [dictionary.doc2bow(text) for text in texts]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "18846\n",
      "Dictionary(134435 unique tokens: [u'3ds2scn', u'diagnositic', u'9l2t', u'l1tbk', u'porkification']...)\n"
     ]
    }
   ],
   "source": [
    "print len(documents)\n",
    "print dictionary"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "topics = []  # list of 100 topics\n",
    "for l in open('/home/devashish/datasets/20NG/topics20NG.txt'):\n",
    "    topics.append([l.split()])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "human_scores = []\n",
    "for l in open('/home/devashish/datasets/20NG/gold20NG.txt'):\n",
    "    human_scores.append(float(l.strip()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Time taken: 0:04:10.235838\n"
     ]
    }
   ],
   "source": [
    "start = datetime.now()\n",
    "u_mass = []\n",
    "flags = []\n",
    "for n, topic in enumerate(topics):\n",
    "    try:\n",
    "        cm = CoherenceModel(topics=topic, corpus=corpus, dictionary=dictionary, coherence='u_mass')\n",
    "        u_mass.append(cm.get_coherence())\n",
    "    except KeyError:\n",
    "        flags.append(n)\n",
    "end = datetime.now()\n",
    "print \"Time taken: %s\" % (end - start)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Time taken: 0:50:03.377507\n"
     ]
    }
   ],
   "source": [
    "start = datetime.now()\n",
    "c_v = []\n",
    "for n, topic in enumerate(topics):\n",
    "    try:\n",
    "        cm = CoherenceModel(topics=topic, texts=texts, dictionary=dictionary, coherence='c_v')\n",
    "        c_v.append(cm.get_coherence())\n",
    "    except KeyError:\n",
    "        pass\n",
    "end = datetime.now()\n",
    "print \"Time taken: %s\" % (end - start)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "final_scores = []\n",
    "for n, score in enumerate(human_scores):\n",
    "    if n not in flags:\n",
    "        final_scores.append(score)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "93 93 93\n"
     ]
    }
   ],
   "source": [
    "print len(u_mass), len(c_v), len(final_scores)\n",
    "# 3 topics have words that are not in the dictionary.This is due to a difference\n",
    "# in preprocessing or because of the absence of ~900 documents"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.554915406168\n",
      "0.616997328993\n"
     ]
    }
   ],
   "source": [
    "print pearsonr(u_mass, final_scores)[0]\n",
    "print pearsonr(c_v, final_scores)[0]"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.11"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}