Yelp Review.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Does Sentiment Analysis work ? Analysis using AFINN and Textblob on Python"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Sentiment Analysis is used to sense the mood of the people based on the text they write. The text could be a tweet, review or anything whose sentiment is to be quantified. The sentiment could be positive or negative. SA(Sentiment Analysis) is difficult to validate because in many scenerios \"ground truth\" is not available. But in the case of Yelp Review Data, both the user review and ratings are available. Therefore rating(1 to 5) value can be used as ground truth information to validate the SA score computed. \n",
    "\n",
    "The data is obtained from [this link from Yelp](https://www.yelp.com/dataset_challenge). The data is downloaded and stored into a Dataframe. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 142,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "from textblob import TextBlob\n",
    "import json\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline  \n",
    "from  matplotlib import style\n",
    "import pandas as pd\n",
    "from itertools import *\n",
    "from matplotlib import rcParams\n",
    "rcParams.update({'figure.autolayout': True})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 143,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "file_name = '/Users/harshkn/Documents/HobbyProjects/YelpReviewData/yelp_dataset_challenge_academic_dataset/yelp_academic_dataset_review.json'\n",
    "# json_data = open(file_name)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 152,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# data = pd.read_json(file_name,lines=True)\n",
    "data = []\n",
    "with open(file_name) as f:\n",
    "    for count,line in islice(enumerate(f), 250000):\n",
    "        data.append(json.loads(line))\n",
    "            \n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Each entry in the file looks as shown below. It has various fields but I'll consider only 'stars' and 'text' fields."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 209,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{u'business_id': u'5UmKMjUEUNdYWqANhGckJw',\n",
       " u'date': u'2012-08-01',\n",
       " u'review_id': u'Ya85v4eqdd6k9Od8HbQjyA',\n",
       " u'stars': 4,\n",
       " u'text': u'Mr Hoagie is an institution. Walking in, it does seem like a throwback to 30 years ago, old fashioned menu board, booths out of the 70s, and a large selection of food. Their speciality is the Italian Hoagie, and it is voted the best in the area year after year. I usually order the burger, while the patties are obviously cooked from frozen, all of the other ingredients are very fresh. Overall, its a good alternative to Subway, which is down the road.',\n",
       " u'type': u'review',\n",
       " u'user_id': u'PUFPaY9KxDAcGqfsorJp3Q',\n",
       " u'votes': {u'cool': 0, u'funny': 0, u'useful': 0}}"
      ]
     },
     "execution_count": 209,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data[0]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 210,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 154,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "df = pd.DataFrame(data)\n",
    "FilteredDf = df[['stars', 'text']]\n",
    "del df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 155,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mr Hoagie is an institution. Walking in, it does seem like a throwback to 30 years ago, old fashioned menu board, booths out of the 70s, and a large selection of food. Their speciality is the Italian Hoagie, and it is voted the best in the area year after year. I usually order the burger, while the patties are obviously cooked from frozen, all of the other ingredients are very fresh. Overall, its a good alternative to Subway, which is down the road.\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "stars    250000\n",
       "text     250000\n",
       "dtype: int64"
      ]
     },
     "execution_count": 155,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Sample entry \n",
    "print FilteredDf['text'][0]\n",
    "FilteredDf.count()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Using AFINN python library\n",
    "AFINN is list of English words with each word assigned a positive or negative integer assigned to it in correspondence with the sentiment it carried. Following words and thier AFINN score from [AFINN repository](https://github.com/fnielsen/afinn). We can see words which carry negative sentiment are assigned negative score and vice-versa. \n",
    "\n",
    " ```\n",
    "await\t-1  \n",
    "awaited\t-1  \n",
    "awaits\t-1  \n",
    "award\t3  \n",
    "awarded\t3  \n",
    "awards\t3  \n",
    "awesome\t4  \n",
    "awful\t-3  \n",
    "awkward\t-2  \n",
    "axe\t-1  \n",
    "axed\t-1  \n",
    "backed\t1  \n",
    "backing\t2  \n",
    "backs\t1  \n",
    "bad\t-3  \n",
    "bad luck\t-2\n",
    "```\n",
    "\n",
    "Sample usage of AFINN is shown below. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 214,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "AFINN score for 'This is utterly excellent!' is  3.0\n"
     ]
    }
   ],
   "source": [
    "from afinn import Afinn\n",
    "afinn = Afinn()\n",
    "print 'AFINN score for \\'This is utterly excellent!\\' is ', afinn.score('This is utterly excellent!')\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "NLTK stopword list is used to remove stop words from the reviews. Before this each review is tokenized into words. And then AFINN lexicon is used to assign sentiment score to each word and average sentiment score for each review is computed. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 221,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "#list of stop words\n",
    "from nltk.corpus import stopwords \n",
    "#to tokenize a sentence into words\n",
    "from nltk.tokenize import wordpunct_tokenize \n",
    "import numpy as np\n",
    "\n",
    "\n",
    "# wordpunct_tokenize(FilteredDf['text'][100])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 157,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "250000"
      ]
     },
     "execution_count": 157,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# To compute average sentiment score for each review\n",
    "stop_words = set(stopwords.words('english'))\n",
    "stop_words.update(['mr','.', ',', '\"', \"'\", '?', '!', ':', ';', '(', ')', '[', ']', '{', '}']) # remove it if you need punctuation \n",
    "unigrams = []\n",
    "afinn_value = []\n",
    "stars = []\n",
    "for ind,review in islice(FilteredDf.iterrows(),250000):\n",
    "    unigrams = ( [i.lower() for i in wordpunct_tokenize(review['text']) if i.lower() not in stop_words])\n",
    "    afinn_value.append(np.mean(list(map(lambda x: afinn.score(str(x.encode('utf-8'))), unigrams))))\n",
    "    stars.append(review['stars'])\n",
    "\n",
    "len(afinn_value)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 158,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1"
      ]
     },
     "execution_count": 158,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pro_stars = pd.DataFrame()\n",
    "pro_stars['stars'] = stars\n",
    "pro_stars['senti_value'] = afinn_value\n",
    "pro_stars['stars'].min()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Later, sentiment score for each star rating is plotted as boxplot as shown below. We can see there is a some correlation between star rating and the computed average sentiment score. But we can see there is lot of variance in the box plots. This is because there are lot of words for which AFINN scores are not available and since we compute average of the total sentence, the score is normalised heavily in many cases. It can be improved by considering only words in AFINN dictionary to compute the average. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 201,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.text.Text at 0x130300e50>"
      ]
     },
     "execution_count": 201,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAr0AAAGdCAYAAAAfYMtzAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nFL0AAADweP8P2PrGNnoI2wwAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1302f34d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "rcParams.update({'font.size': 15})\n",
    "plt.style.use('seaborn-white')\n",
    "ax = pro_stars.boxplot(by=['stars'], figsize=(10,6))\n",
    "ax.get_figure().suptitle(\"\")\n",
    "ax.set_title('Average sentiment score and Star rating. - Using AFINN')\n",
    "ax.set_xlabel('stars')\n",
    "ax.set_ylabel('Average sentiment score')\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Using Textblob\n",
    "\n",
    "Now, we use textblob, a text processing library which also has sentiment analysis api along with many others. Sample usage of Textblob is shown below. Later we use each review in full as an input to textblob api and compute the sentiment score. We can see from the plot below that textblob has better correlation with the star ratings than from the one obtained using AFINN. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 179,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "import textblob as tb"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 205,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mr Hoagie is an institution. Walking in, it does seem like a throwback to 30 years ago, old fashioned menu board, booths out of the 70s, and a large selection of food. Their speciality is the Italian Hoagie, and it is voted the best in the area year after year. I usually order the burger, while the patties are obviously cooked from frozen, all of the other ingredients are very fresh. Overall, its a good alternative to Subway, which is down the road.\n",
      "4\n",
      "\n",
      "Sentiment(polarity=0.1703391053391053, subjectivity=0.3265873015873016)\n",
      "\n"
     ]
    }
   ],
   "source": [
    "#Demo \n",
    "print FilteredDf['text'][0]\n",
    "print FilteredDf['stars'][0]\n",
    "details = TextBlob(FilteredDf['text'][0])\n",
    "print\n",
    "print details.sentiment\n",
    "print\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 202,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "tb_value = []\n",
    "stars_tb = []\n",
    "for ind,review in islice(FilteredDf.iterrows(),250000):\n",
    "    details = TextBlob(review['text'])\n",
    "    tb_value.append(details.sentiment.polarity)\n",
    "    stars_tb.append(review['stars'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 203,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "pro_stars_tb = pd.DataFrame()\n",
    "pro_stars_tb['stars'] = stars_tb\n",
    "pro_stars_tb['senti_value'] = tb_value\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 204,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.text.Text at 0x130314790>"
      ]
     },
     "execution_count": 204,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAr0AAAGdCAYAAAAfYMtzAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n7/8Bg1+MGTtKZA0AAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x145770b50>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "ax = pro_stars_tb.boxplot(by=['stars'], figsize=(10,6))\n",
    "ax.get_figure().suptitle(\"\")\n",
    "ax.set_title('Average sentiment score and Star rating - Using Textblob.')\n",
    "ax.set_xlabel('stars')\n",
    "ax.set_ylabel('Average sentiment score')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This post is inspired by David Robinson's [post](http://varianceexplained.org/r/yelp-sentiment/) on sentiment analysis with Yelp review data. Since the post referred uses R, I wanted to try it using Python. "
   ]
  }
 ],
 "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": "2.7.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}