sentimmeal_analysis_Hotel_reviews.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Sentimental Analsyis : Data Science and NLP for Customer Reviews\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[nltk_data] Downloading package vader_lexicon to\n",
      "[nltk_data]     C:\\Users\\Sundeep\\AppData\\Roaming\\nltk_data...\n",
      "[nltk_data]   Package vader_lexicon is already up-to-date!\n",
      "[nltk_data] Downloading package stopwords to\n",
      "[nltk_data]     C:\\Users\\Sundeep\\AppData\\Roaming\\nltk_data...\n",
      "[nltk_data]   Package stopwords is already up-to-date!\n",
      "[nltk_data] Downloading package averaged_perceptron_tagger to\n",
      "[nltk_data]     C:\\Users\\Sundeep\\AppData\\Roaming\\nltk_data...\n",
      "[nltk_data]   Package averaged_perceptron_tagger is already up-to-\n",
      "[nltk_data]       date!\n",
      "[nltk_data] Downloading package wordnet to\n",
      "[nltk_data]     C:\\Users\\Sundeep\\AppData\\Roaming\\nltk_data...\n",
      "[nltk_data]   Package wordnet is already up-to-date!\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\Sundeep\\Anaconda3\\lib\\site-packages\\nltk\\twitter\\__init__.py:20: UserWarning: The twython library has not been installed. Some functionality from the twitter package will not be available.\n",
      "  warnings.warn(\"The twython library has not been installed. \"\n"
     ]
    }
   ],
   "source": [
    "#Plot\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "%matplotlib inline\n",
    "\n",
    "#Data Packages\n",
    "import math\n",
    "import pandas\n",
    "import numpy as np\n",
    "\n",
    "#Progress bar\n",
    "from tqdm import tqdm\n",
    "\n",
    "#Counter\n",
    "from collections import Counter\n",
    "\n",
    "#Operation\n",
    "import operator\n",
    "\n",
    "#Natural Language Processing Packages\n",
    "import re\n",
    "import nltk\n",
    "\n",
    "## Download Resources\n",
    "nltk.download(\"vader_lexicon\")\n",
    "nltk.download(\"stopwords\")\n",
    "nltk.download(\"averaged_perceptron_tagger\")\n",
    "nltk.download(\"wordnet\")\n",
    "\n",
    "from nltk.sentiment import SentimentAnalyzer\n",
    "from nltk.sentiment.vader import SentimentIntensityAnalyzer\n",
    "from nltk.sentiment.util import *\n",
    "from nltk import tokenize\n",
    "from nltk.corpus import stopwords\n",
    "from nltk.tag import PerceptronTagger\n",
    "from nltk.data import find\n",
    "\n",
    "## Machine Learning\n",
    "import sklearn\n",
    "import sklearn.metrics as metrics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Read in from pandas\n",
    "hotelDf = pandas.read_csv('reviews_calgary.csv')\n",
    "hotelDf.columns=['filePath','hotelName','reviewColumn','ratingScore','groundTruth']\n",
    "hotelDf = hotelDf.drop(columns = ['filePath'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>hotelName</th>\n",
       "      <th>reviewColumn</th>\n",
       "      <th>ratingScore</th>\n",
       "      <th>groundTruth</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Calgary Airport Marriott In-Terminal Hotel</td>\n",
       "      <td>\"We were excited to stay at the new Marriott i...</td>\n",
       "      <td>5</td>\n",
       "      <td>positive</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Calgary Airport Marriott In-Terminal Hotel</td>\n",
       "      <td>\"We were not able to make our Victoria connect...</td>\n",
       "      <td>5</td>\n",
       "      <td>positive</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Calgary Airport Marriott In-Terminal Hotel</td>\n",
       "      <td>\"Rooms are very modern looking with great king...</td>\n",
       "      <td>5</td>\n",
       "      <td>positive</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Calgary Airport Marriott In-Terminal Hotel</td>\n",
       "      <td>\"We stayed just one night at the Marriott and ...</td>\n",
       "      <td>4</td>\n",
       "      <td>positive</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Calgary Airport Marriott In-Terminal Hotel</td>\n",
       "      <td>\"The Marriot is across from the old Delta. Bot...</td>\n",
       "      <td>5</td>\n",
       "      <td>positive</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                    hotelName  \\\n",
       "0  Calgary Airport Marriott In-Terminal Hotel   \n",
       "1  Calgary Airport Marriott In-Terminal Hotel   \n",
       "2  Calgary Airport Marriott In-Terminal Hotel   \n",
       "3  Calgary Airport Marriott In-Terminal Hotel   \n",
       "4  Calgary Airport Marriott In-Terminal Hotel   \n",
       "\n",
       "                                        reviewColumn  ratingScore groundTruth  \n",
       "0  \"We were excited to stay at the new Marriott i...            5    positive  \n",
       "1  \"We were not able to make our Victoria connect...            5    positive  \n",
       "2  \"Rooms are very modern looking with great king...            5    positive  \n",
       "3  \"We stayed just one night at the Marriott and ...            4    positive  \n",
       "4  \"The Marriot is across from the old Delta. Bot...            5    positive  "
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "hotelDf.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "positive    9434\n",
       "negative    3249\n",
       "Name: groundTruth, dtype: int64"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "hotelDf['groundTruth'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 12683 entries, 0 to 12682\n",
      "Data columns (total 4 columns):\n",
      "hotelName       12683 non-null object\n",
      "reviewColumn    12683 non-null object\n",
      "ratingScore     12683 non-null int64\n",
      "groundTruth     12683 non-null object\n",
      "dtypes: int64(1), object(3)\n",
      "memory usage: 396.4+ KB\n"
     ]
    }
   ],
   "source": [
    "hotelDf.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "review_counts = hotelDf['hotelName'].value_counts()\n",
    "review_counts = dict(review_counts)\n",
    "hotelDf['review_counts'] = hotelDf['hotelName'].apply(lambda x: review_counts[x])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Use vader to evaluated sentiment of reviews\n",
    "def evalSentences(sentences):\n",
    "    #Instantiate an instance to access SentimentIntensityAnalyzer class\n",
    "    sid = SentimentIntensityAnalyzer()\n",
    "    pdlist = []\n",
    "\n",
    "    for sentence in sentences:\n",
    "        ss = sid.polarity_scores(sentence)\n",
    "        pdlist.append(ss['compound'])\n",
    "        \n",
    "    return pdlist\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "reviews = hotelDf['reviewColumn'].values\n",
    "Vader_sentiment_scores = evalSentences(reviews)\n",
    "reviews = hotelDf['reviewColumn'].values\n",
    "hotelDf['Vader_score'] = Vader_sentiment_scores"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>hotelName</th>\n",
       "      <th>reviewColumn</th>\n",
       "      <th>ratingScore</th>\n",
       "      <th>groundTruth</th>\n",
       "      <th>review_counts</th>\n",
       "      <th>Vader_score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Calgary Airport Marriott In-Terminal Hotel</td>\n",
       "      <td>\"We were excited to stay at the new Marriott i...</td>\n",
       "      <td>5</td>\n",
       "      <td>positive</td>\n",
       "      <td>209</td>\n",
       "      <td>0.9830</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Calgary Airport Marriott In-Terminal Hotel</td>\n",
       "      <td>\"We were not able to make our Victoria connect...</td>\n",
       "      <td>5</td>\n",
       "      <td>positive</td>\n",
       "      <td>209</td>\n",
       "      <td>0.9412</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Calgary Airport Marriott In-Terminal Hotel</td>\n",
       "      <td>\"Rooms are very modern looking with great king...</td>\n",
       "      <td>5</td>\n",
       "      <td>positive</td>\n",
       "      <td>209</td>\n",
       "      <td>0.6409</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Calgary Airport Marriott In-Terminal Hotel</td>\n",
       "      <td>\"We stayed just one night at the Marriott and ...</td>\n",
       "      <td>4</td>\n",
       "      <td>positive</td>\n",
       "      <td>209</td>\n",
       "      <td>0.4957</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Calgary Airport Marriott In-Terminal Hotel</td>\n",
       "      <td>\"The Marriot is across from the old Delta. Bot...</td>\n",
       "      <td>5</td>\n",
       "      <td>positive</td>\n",
       "      <td>209</td>\n",
       "      <td>0.9735</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                    hotelName  \\\n",
       "0  Calgary Airport Marriott In-Terminal Hotel   \n",
       "1  Calgary Airport Marriott In-Terminal Hotel   \n",
       "2  Calgary Airport Marriott In-Terminal Hotel   \n",
       "3  Calgary Airport Marriott In-Terminal Hotel   \n",
       "4  Calgary Airport Marriott In-Terminal Hotel   \n",
       "\n",
       "                                        reviewColumn  ratingScore groundTruth  \\\n",
       "0  \"We were excited to stay at the new Marriott i...            5    positive   \n",
       "1  \"We were not able to make our Victoria connect...            5    positive   \n",
       "2  \"Rooms are very modern looking with great king...            5    positive   \n",
       "3  \"We stayed just one night at the Marriott and ...            4    positive   \n",
       "4  \"The Marriot is across from the old Delta. Bot...            5    positive   \n",
       "\n",
       "   review_counts  Vader_score  \n",
       "0            209       0.9830  \n",
       "1            209       0.9412  \n",
       "2            209       0.6409  \n",
       "3            209       0.4957  \n",
       "4            209       0.9735  "
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "hotelDf.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Calgary Marriott Downtown Hotel       643\n",
       "The Westin Calgary                    622\n",
       "Sheraton Suites Calgary Eau Claire    536\n",
       "The Fairmont Palliser                 533\n",
       "Hyatt Regency Calgary                 492\n",
       "Name: hotelName, dtype: int64"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "hotelDf.hotelName.value_counts().head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "hotels_avg_data = hotelDf.groupby('hotelName').mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>ratingScore</th>\n",
       "      <th>review_counts</th>\n",
       "      <th>Vader_score</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>hotelName</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>11th Street Lodging B&amp;amp;B</th>\n",
       "      <td>4.500000</td>\n",
       "      <td>12.0</td>\n",
       "      <td>0.807492</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Acclaim Hotel Calgary Airport</th>\n",
       "      <td>4.072165</td>\n",
       "      <td>97.0</td>\n",
       "      <td>0.741924</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Aloft Calgary University</th>\n",
       "      <td>4.214689</td>\n",
       "      <td>177.0</td>\n",
       "      <td>0.812512</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Along River Ridge</th>\n",
       "      <td>4.875000</td>\n",
       "      <td>56.0</td>\n",
       "      <td>0.945861</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Ambassador Motor Inn</th>\n",
       "      <td>2.562500</td>\n",
       "      <td>16.0</td>\n",
       "      <td>0.248481</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                               ratingScore  review_counts  Vader_score\n",
       "hotelName                                                             \n",
       "11th Street Lodging B&amp;B       4.500000           12.0     0.807492\n",
       "Acclaim Hotel Calgary Airport     4.072165           97.0     0.741924\n",
       "Aloft Calgary University          4.214689          177.0     0.812512\n",
       "Along River Ridge                 4.875000           56.0     0.945861\n",
       "Ambassador Motor Inn              2.562500           16.0     0.248481"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "hotels_avg_data.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Top 5 Hotels "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>ratingScore</th>\n",
       "      <th>review_counts</th>\n",
       "      <th>Vader_score</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>hotelName</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Brother Li Homestay</th>\n",
       "      <td>5.000000</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.957100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Bonaventure Bed and Breakfast</th>\n",
       "      <td>4.937500</td>\n",
       "      <td>16.0</td>\n",
       "      <td>0.952269</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Novel Bed and Breakfast</th>\n",
       "      <td>4.923077</td>\n",
       "      <td>13.0</td>\n",
       "      <td>0.947500</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Along River Ridge</th>\n",
       "      <td>4.875000</td>\n",
       "      <td>56.0</td>\n",
       "      <td>0.945861</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>English Bed, Breakfast and Day Spa</th>\n",
       "      <td>4.848485</td>\n",
       "      <td>33.0</td>\n",
       "      <td>0.886221</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                    ratingScore  review_counts  Vader_score\n",
       "hotelName                                                                  \n",
       "Brother Li Homestay                    5.000000            2.0     0.957100\n",
       "Bonaventure Bed and Breakfast          4.937500           16.0     0.952269\n",
       "Novel Bed and Breakfast                4.923077           13.0     0.947500\n",
       "Along River Ridge                      4.875000           56.0     0.945861\n",
       "English Bed, Breakfast and Day Spa     4.848485           33.0     0.886221"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "hotels_avg_data.sort_values(by=['ratingScore'], ascending=False).head(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>ratingScore</th>\n",
       "      <th>review_counts</th>\n",
       "      <th>Vader_score</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>hotelName</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Regal Suites - Riverfront</th>\n",
       "      <td>3.750000</td>\n",
       "      <td>4.0</td>\n",
       "      <td>0.982125</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>River Lee Executive Bed &amp;amp; Breakfast</th>\n",
       "      <td>4.636364</td>\n",
       "      <td>11.0</td>\n",
       "      <td>0.964745</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Brother Li Homestay</th>\n",
       "      <td>5.000000</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.957100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Bonaventure Bed and Breakfast</th>\n",
       "      <td>4.937500</td>\n",
       "      <td>16.0</td>\n",
       "      <td>0.952269</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Novel Bed and Breakfast</th>\n",
       "      <td>4.923077</td>\n",
       "      <td>13.0</td>\n",
       "      <td>0.947500</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                         ratingScore  review_counts  \\\n",
       "hotelName                                                             \n",
       "Regal Suites - Riverfront                   3.750000            4.0   \n",
       "River Lee Executive Bed &amp; Breakfast     4.636364           11.0   \n",
       "Brother Li Homestay                         5.000000            2.0   \n",
       "Bonaventure Bed and Breakfast               4.937500           16.0   \n",
       "Novel Bed and Breakfast                     4.923077           13.0   \n",
       "\n",
       "                                         Vader_score  \n",
       "hotelName                                             \n",
       "Regal Suites - Riverfront                   0.982125  \n",
       "River Lee Executive Bed &amp; Breakfast     0.964745  \n",
       "Brother Li Homestay                         0.957100  \n",
       "Bonaventure Bed and Breakfast               0.952269  \n",
       "Novel Bed and Breakfast                     0.947500  "
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "hotels_avg_data.sort_values(by=['Vader_score'], ascending=False).head(5)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Bottom-5 Hotels"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>ratingScore</th>\n",
       "      <th>review_counts</th>\n",
       "      <th>Vader_score</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>hotelName</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Red Carpet Inn</th>\n",
       "      <td>2.052632</td>\n",
       "      <td>19.0</td>\n",
       "      <td>0.203258</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Ramada Limited Calgary Northwest</th>\n",
       "      <td>1.809211</td>\n",
       "      <td>152.0</td>\n",
       "      <td>-0.046259</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Town &amp;amp; Country Motor Hotel</th>\n",
       "      <td>1.300000</td>\n",
       "      <td>10.0</td>\n",
       "      <td>-0.112840</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Olympia Motel</th>\n",
       "      <td>1.090909</td>\n",
       "      <td>11.0</td>\n",
       "      <td>-0.025009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>O Stays Waterfront</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>6.0</td>\n",
       "      <td>0.173583</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                  ratingScore  review_counts  Vader_score\n",
       "hotelName                                                                \n",
       "Red Carpet Inn                       2.052632           19.0     0.203258\n",
       "Ramada Limited Calgary Northwest     1.809211          152.0    -0.046259\n",
       "Town &amp; Country Motor Hotel       1.300000           10.0    -0.112840\n",
       "Olympia Motel                        1.090909           11.0    -0.025009\n",
       "O Stays Waterfront                   1.000000            6.0     0.173583"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "hotels_avg_data.sort_values(by=['ratingScore'], ascending=False).tail()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>ratingScore</th>\n",
       "      <th>review_counts</th>\n",
       "      <th>Vader_score</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>hotelName</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>O Stays Waterfront</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>6.0</td>\n",
       "      <td>0.173583</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Olympia Motel</th>\n",
       "      <td>1.090909</td>\n",
       "      <td>11.0</td>\n",
       "      <td>-0.025009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Canadas Best Value Inn Chinook Station</th>\n",
       "      <td>2.079208</td>\n",
       "      <td>202.0</td>\n",
       "      <td>-0.045217</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Ramada Limited Calgary Northwest</th>\n",
       "      <td>1.809211</td>\n",
       "      <td>152.0</td>\n",
       "      <td>-0.046259</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Town &amp;amp; Country Motor Hotel</th>\n",
       "      <td>1.300000</td>\n",
       "      <td>10.0</td>\n",
       "      <td>-0.112840</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                        ratingScore  review_counts  \\\n",
       "hotelName                                                            \n",
       "O Stays Waterfront                         1.000000            6.0   \n",
       "Olympia Motel                              1.090909           11.0   \n",
       "Canadas Best Value Inn Chinook Station     2.079208          202.0   \n",
       "Ramada Limited Calgary Northwest           1.809211          152.0   \n",
       "Town &amp; Country Motor Hotel             1.300000           10.0   \n",
       "\n",
       "                                        Vader_score  \n",
       "hotelName                                            \n",
       "O Stays Waterfront                         0.173583  \n",
       "Olympia Motel                             -0.025009  \n",
       "Canadas Best Value Inn Chinook Station    -0.045217  \n",
       "Ramada Limited Calgary Northwest          -0.046259  \n",
       "Town &amp; Country Motor Hotel            -0.112840  "
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "hotels_avg_data.sort_values(by=['Vader_score'], ascending=False).tail()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Comparision for top-5 and bottom-5 for both ranking methods:"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can see that most of the top-5 and bottom-5 hotels are matching for vader & rating score. Although there is differnce in the rankings but top & bottom 5 hotels are almost same. (3 Hotels matching in Top Hotels & 4 Hotels matching in Bottom hotels)\n",
    "\n",
    "Also, in both the cases while ranking hotels by **average rating scores**, we can observe a clear difference in rankings for the hotels with less number of reviews which is casuing slight difference in rakings for rating & vader score methods. (Hotels with less number of reviews rating score ranking is dominating)\n",
    "\n",
    "Overall we can say that Vader using lexicon and rule-based sentiment analysis is making good predictons for very positive and very negative reviews."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Q2. Frequency Analysis"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Note: You may want to use an NLTK tokenizer instead of a regular expression in the following\n",
    "def dataFrameTransformation(hotelDf,k=500):\n",
    "    reviews = hotelDf['reviewColumn'].values\n",
    "    \n",
    "    from nltk.corpus import stopwords\n",
    "    stop = set(stopwords.words('english'))\n",
    "    \n",
    "    # Top-k frequent terms\n",
    "    counter = Counter()\n",
    "    for review in reviews:\n",
    "            counter.update([word.lower() \n",
    "                            for word \n",
    "                            in re.findall(r'\\w+', review) \n",
    "                            if word.lower() not in stop and len(word) > 2])\n",
    "    topk = counter.most_common(k)        \n",
    "    \n",
    "    #Find out if a particular review has the word from topk list\n",
    "    freqReview = []\n",
    "    for i in range(len(reviews)):\n",
    "        tempCounter = Counter([word.lower() for word in re.findall(r'\\w+',reviews[i])])\n",
    "        topkinReview = [1 if tempCounter[word] > 0 else 0 for (word,wordCount) in topk]\n",
    "        freqReview.append(topkinReview)\n",
    "        \n",
    "        \n",
    "    #Prepare freqReviewDf\n",
    "    freqReviewDf = pandas.DataFrame(freqReview)\n",
    "    dfName = []\n",
    "    for c in topk:\n",
    "        dfName.append(c[0])\n",
    "    freqReviewDf.columns = dfName\n",
    "#     finalreviewDf = reviewDF.join(freqReviewDf)\n",
    "    finaldf = hotelDf[['hotelName','reviewColumn','ratingScore','groundTruth','Vader_score']].join(freqReviewDf)\n",
    "    return topk, finaldf"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[('hotel', 15062),\n",
       " ('room', 13731),\n",
       " ('staff', 7472),\n",
       " ('stay', 7142),\n",
       " ('great', 6484),\n",
       " ('good', 6184),\n",
       " ('clean', 5309),\n",
       " ('calgary', 5285),\n",
       " ('breakfast', 5273),\n",
       " ('would', 4921)]"
      ]
     },
     "execution_count": 96,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "topk, finaldf = dataFrameTransformation(hotelDf, k=500)\n",
    "topk[:10]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 97,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>hotelName</th>\n",
       "      <th>reviewColumn</th>\n",
       "      <th>ratingScore</th>\n",
       "      <th>groundTruth</th>\n",
       "      <th>Vader_score</th>\n",
       "      <th>hotel</th>\n",
       "      <th>room</th>\n",
       "      <th>staff</th>\n",
       "      <th>stay</th>\n",
       "      <th>great</th>\n",
       "      <th>...</th>\n",
       "      <th>variety</th>\n",
       "      <th>using</th>\n",
       "      <th>real</th>\n",
       "      <th>elevators</th>\n",
       "      <th>maintained</th>\n",
       "      <th>adequate</th>\n",
       "      <th>beyond</th>\n",
       "      <th>choices</th>\n",
       "      <th>nthere</th>\n",
       "      <th>blocks</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Calgary Airport Marriott In-Terminal Hotel</td>\n",
       "      <td>\"We were excited to stay at the new Marriott i...</td>\n",
       "      <td>5</td>\n",
       "      <td>positive</td>\n",
       "      <td>0.9830</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Calgary Airport Marriott In-Terminal Hotel</td>\n",
       "      <td>\"We were not able to make our Victoria connect...</td>\n",
       "      <td>5</td>\n",
       "      <td>positive</td>\n",
       "      <td>0.9412</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Calgary Airport Marriott In-Terminal Hotel</td>\n",
       "      <td>\"Rooms are very modern looking with great king...</td>\n",
       "      <td>5</td>\n",
       "      <td>positive</td>\n",
       "      <td>0.6409</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>3 rows × 505 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                    hotelName  \\\n",
       "0  Calgary Airport Marriott In-Terminal Hotel   \n",
       "1  Calgary Airport Marriott In-Terminal Hotel   \n",
       "2  Calgary Airport Marriott In-Terminal Hotel   \n",
       "\n",
       "                                        reviewColumn  ratingScore groundTruth  \\\n",
       "0  \"We were excited to stay at the new Marriott i...            5    positive   \n",
       "1  \"We were not able to make our Victoria connect...            5    positive   \n",
       "2  \"Rooms are very modern looking with great king...            5    positive   \n",
       "\n",
       "   Vader_score  hotel  room  staff  stay  great   ...    variety  using  real  \\\n",
       "0       0.9830      1     1      0     1      1   ...          0      0     0   \n",
       "1       0.9412      1     0      1     0      0   ...          0      0     0   \n",
       "2       0.6409      0     1      0     1      1   ...          0      0     0   \n",
       "\n",
       "   elevators  maintained  adequate  beyond  choices  nthere  blocks  \n",
       "0          0           0         0       0        0       0       0  \n",
       "1          0           0         0       1        0       0       0  \n",
       "2          1           0         0       0        0       0       0  \n",
       "\n",
       "[3 rows x 505 columns]"
      ]
     },
     "execution_count": 97,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "finaldf.head(3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 126,
   "metadata": {},
   "outputs": [],
   "source": [
    "def getTopK(df, k, label_value, label_column='groundTruth', operation=operator.eq, value_column='reviewColumn'):\n",
    "    from nltk.corpus import stopwords\n",
    "    stop = set(stopwords.words('english'))\n",
    "    #Add possible Stop Words for Hotel Reviews\n",
    "    stop.add('hotel')\n",
    "    stop.add('room')\n",
    "    stop.add('rooms')\n",
    "    stop.add('stay')\n",
    "    stop.add('would')\n",
    "    stop.add('calgary')\n",
    "#     stop.add('told')\n",
    "#     stop.add('said')\n",
    "#     stop.add('called')\n",
    "    counter = Counter()\n",
    "    for review in df.loc[operation(df[label_column],label_value)][value_column]:\n",
    "            counter.update([word.lower() \n",
    "                            for word \n",
    "                            in re.findall(r'\\w+', review) \n",
    "                            if word.lower() not in stop and len(word) > 2])\n",
    "    topk = counter.most_common(k)\n",
    "    return topk"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Staff is not considerered as a stop word since it helps us to identify how many customers are mentiioning about staff in positive and negative reviews. This gives us their overall opinion on the staff."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 127,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[('staff', 5904),\n",
       " ('great', 5787),\n",
       " ('good', 4774),\n",
       " ('clean', 4153),\n",
       " ('breakfast', 4021),\n",
       " ('nice', 3588),\n",
       " ('friendly', 3034),\n",
       " ('stayed', 3003),\n",
       " ('comfortable', 2911),\n",
       " ('service', 2899),\n",
       " ('location', 2647),\n",
       " ('well', 2592),\n",
       " ('one', 2413),\n",
       " ('downtown', 2062),\n",
       " ('night', 2053),\n",
       " ('helpful', 2028),\n",
       " ('pool', 1949),\n",
       " ('also', 1890),\n",
       " ('excellent', 1855),\n",
       " ('time', 1806),\n",
       " ('restaurant', 1779),\n",
       " ('nthe', 1758),\n",
       " ('bed', 1750),\n",
       " ('desk', 1718),\n",
       " ('area', 1714),\n",
       " ('food', 1688),\n",
       " ('parking', 1640),\n",
       " ('airport', 1580),\n",
       " ('free', 1567),\n",
       " ('close', 1544),\n",
       " ('front', 1542),\n",
       " ('get', 1542),\n",
       " ('really', 1530),\n",
       " ('check', 1474),\n",
       " ('place', 1442),\n",
       " ('restaurants', 1297),\n",
       " ('recommend', 1290),\n",
       " ('like', 1283),\n",
       " ('beds', 1267),\n",
       " ('could', 1235),\n",
       " ('quot', 1213),\n",
       " ('day', 1201),\n",
       " ('back', 1195),\n",
       " ('always', 1156),\n",
       " ('hot', 1150),\n",
       " ('definitely', 1136),\n",
       " ('best', 1080),\n",
       " ('business', 1069),\n",
       " ('floor', 1053),\n",
       " ('walk', 1052)]"
      ]
     },
     "execution_count": 127,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "topkGroundPos = getTopK(df=finaldf, k=50, label_value='positive')\n",
    "topkGroundPos"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 128,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[('staff', 1568),\n",
       " ('one', 1534),\n",
       " ('night', 1506),\n",
       " ('good', 1410),\n",
       " ('breakfast', 1252),\n",
       " ('desk', 1222),\n",
       " ('quot', 1164),\n",
       " ('clean', 1156),\n",
       " ('get', 1142),\n",
       " ('stayed', 1120),\n",
       " ('front', 1120),\n",
       " ('bed', 961),\n",
       " ('could', 947),\n",
       " ('service', 883),\n",
       " ('nice', 880),\n",
       " ('nthe', 856),\n",
       " ('like', 854),\n",
       " ('time', 851),\n",
       " ('place', 779),\n",
       " ('check', 777),\n",
       " ('even', 769),\n",
       " ('floor', 717),\n",
       " ('back', 703),\n",
       " ('great', 697),\n",
       " ('parking', 696),\n",
       " ('bathroom', 687),\n",
       " ('location', 678),\n",
       " ('day', 668),\n",
       " ('also', 649),\n",
       " ('well', 626),\n",
       " ('friendly', 609),\n",
       " ('area', 599),\n",
       " ('small', 590),\n",
       " ('door', 587),\n",
       " ('pool', 583),\n",
       " ('next', 580),\n",
       " ('got', 572),\n",
       " ('booked', 570),\n",
       " ('two', 556),\n",
       " ('first', 539),\n",
       " ('never', 537),\n",
       " ('people', 536),\n",
       " ('really', 534),\n",
       " ('hot', 519),\n",
       " ('comfortable', 511),\n",
       " ('work', 505),\n",
       " ('morning', 496),\n",
       " ('asked', 494),\n",
       " ('food', 492),\n",
       " ('shower', 491)]"
      ]
     },
     "execution_count": 128,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "topkGroundNeg = getTopK(df=finaldf, k=50, label_value='negative')\n",
    "topkGroundNeg"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Insights about Top Ranked Positive & Negative review words:"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In positive reviews we can find some of the top frequent words as - Great/Good/clean/nice/comfortable/excellent etc., as expected. \n",
    "\n",
    "But, Interestingly even in the negative reviews we can find same words in the top list - (Good/clean/nice etc.,). This might be because customers are using negation of these words like (Not Good/wasn't very good/Not clean etc.,) which cannot captured if we are doing single word analysis. (Vader can handle these because of the lexicon and rule based analysis) \n",
    "\n",
    "Also, in both the cases (positive and negative) we can find some other most frequent words like Staff/breakfast/parking/service/food etc., which shows us what customers are more concerned about. By performing similar analysis for an individual hotel we can uderstand about these things in detail (whether customers are positive or negative about these different things)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 129,
   "metadata": {},
   "outputs": [],
   "source": [
    "tagger = PerceptronTagger()\n",
    "\n",
    "pos_tag = tagger.tag"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 130,
   "metadata": {},
   "outputs": [],
   "source": [
    "# This grammar is described in the paper by S. N. Kim,\n",
    "# T. Baldwin, and M.-Y. Kan.\n",
    "# Evaluating n-gram based evaluation metrics for automatic\n",
    "# keyphrase extraction.\n",
    "# Technical report, University of Melbourne, Melbourne 2010.\n",
    "\n",
    "# We will consider Noun Phrase Chunking and we search for chunks corresponding to an individual noun phrase.\n",
    "# Rule states that whenever the chunk finds an Proper Noun (NN) followed by any number of adjectives (JJ) and \n",
    "# then terminated with a noun (NN) then the Noun Phrase(NP) chunk should be formed\n",
    "grammar = r\"\"\"\n",
    "    NBAR:\n",
    "        {<NN.*|JJ>*<NN.*>}  # Nouns and Adjectives, terminated with Nouns\n",
    "        \n",
    "    NP:\n",
    "        {<NBAR>}\n",
    "        {<NBAR><IN><NBAR>}  # Above, connected with in/of/etc...\n",
    "\"\"\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 131,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Create phrase tree\n",
    "# Chunking is a process of extracting phrases from unstructured text.\n",
    "# Chunking works on top of POS tagging, it uses pos-tags as input and provides chunks as output.\n",
    "# Chunking is very important when you want to extract information from text such as Locations, Person Names etc.\n",
    "chunker = nltk.RegexpParser(grammar)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 132,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Noun Phrase Extraction Support Functions\n",
    "from nltk.corpus import stopwords\n",
    "stopwords = stopwords.words('english')\n",
    "lemmatizer = nltk.WordNetLemmatizer()\n",
    "stemmer = nltk.stem.porter.PorterStemmer()\n",
    "\n",
    "# generator, generate leaves one by one\n",
    "def leaves(tree):\n",
    "    \"\"\"Finds NP (nounphrase) leaf nodes of a chunk tree.\"\"\"\n",
    "    for subtree in tree.subtrees(filter = lambda t: t.label()=='NP' or t.label()=='JJ' or t.label()=='RB'):\n",
    "        yield subtree.leaves()\n",
    "\n",
    "# stemming, lematizing, lower case... \n",
    "def normalise(word):\n",
    "    \"\"\"Normalises words to lowercase and stems and lemmatizes it.\"\"\"\n",
    "    word = word.lower()\n",
    "    word = stemmer.stem(word)\n",
    "    word = lemmatizer.lemmatize(word)\n",
    "    return word\n",
    "\n",
    "# stop-words and length control\n",
    "def acceptable_word(word):\n",
    "    \"\"\"Checks conditions for acceptable word: length, stopword.\"\"\"\n",
    "    accepted = bool(2 <= len(word) <= 40\n",
    "        and word.lower() not in stopwords)\n",
    "    return accepted\n",
    "\n",
    "# generator, create item once a time\n",
    "def get_terms(tree):\n",
    "    for leaf in leaves(tree):\n",
    "        term = [normalise(w) for w,t in leaf if acceptable_word(w) ]\n",
    "        # Phrase only\n",
    "        if len(term)>1:\n",
    "            yield term"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 133,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Flatten phrase lists to get tokens for analysis\n",
    "def flatten(npTokenList):\n",
    "    finalList =[]\n",
    "    for phrase in npTokenList:\n",
    "        token = ''\n",
    "        for word in phrase:\n",
    "            token += word + ' '\n",
    "        finalList.append(token.rstrip())\n",
    "    return finalList"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 134,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Revise the previous dataframe transform function...\n",
    "def newDataFrameTransformation(hotelDf, k=500):\n",
    "    reviews = hotelDf['reviewColumn'].values\n",
    "    \n",
    "    # Top-k frequent terms\n",
    "    counter = Counter()\n",
    "    for review in reviews:\n",
    "            counter.update(flatten([word\n",
    "                            for word \n",
    "                            in get_terms(chunker.parse(pos_tag(re.findall(r'\\w+', review)))) \n",
    "                            ]))\n",
    "    topk = counter.most_common(k)        \n",
    "    \n",
    "    #Find out if a particular review has the word from topk list\n",
    "    freqReview = []\n",
    "    for i in range(len(reviews)):\n",
    "        tempCounter = Counter(flatten([word \n",
    "                                       for word \n",
    "                                       in get_terms(chunker.parse(pos_tag(re.findall(r'\\w+',reviews[i]))))]))\n",
    "        topkinReview = [1 if tempCounter[word] > 0 else 0 for (word,wordCount) in topk]\n",
    "        freqReview.append(topkinReview)\n",
    "        \n",
    "        \n",
    "    #Prepare freqReviewDf\n",
    "    freqReviewDf = pandas.DataFrame(freqReview)\n",
    "    dfName = []\n",
    "    for c in topk:\n",
    "        dfName.append(c[0])\n",
    "    freqReviewDf.columns = dfName\n",
    "    freqReviewDf.columns = dfName\n",
    "    finaldf = hotelDf[['hotelName','reviewColumn','ratingScore','groundTruth','Vader_score']].join(freqReviewDf)\n",
    "    return topk, finaldf"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 135,
   "metadata": {},
   "outputs": [],
   "source": [
    "topk_phrase, finaldf_phrase = newDataFrameTransformation(hotelDf)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 136,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>hotelName</th>\n",
       "      <th>reviewColumn</th>\n",
       "      <th>ratingScore</th>\n",
       "      <th>groundTruth</th>\n",
       "      <th>Vader_score</th>\n",
       "      <th>front desk</th>\n",
       "      <th>hot tub</th>\n",
       "      <th>downtown calgari</th>\n",
       "      <th>great locat</th>\n",
       "      <th>front desk staff</th>\n",
       "      <th>...</th>\n",
       "      <th>busi center</th>\n",
       "      <th>nearbi restaur</th>\n",
       "      <th>hair dryer</th>\n",
       "      <th>tran canada highway</th>\n",
       "      <th>waffl maker</th>\n",
       "      <th>room price</th>\n",
       "      <th>riversid inn</th>\n",
       "      <th>larg walk</th>\n",
       "      <th>nice clean</th>\n",
       "      <th>numer time</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Calgary Airport Marriott In-Terminal Hotel</td>\n",
       "      <td>\"We were excited to stay at the new Marriott i...</td>\n",
       "      <td>5</td>\n",
       "      <td>positive</td>\n",
       "      <td>0.9830</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Calgary Airport Marriott In-Terminal Hotel</td>\n",
       "      <td>\"We were not able to make our Victoria connect...</td>\n",
       "      <td>5</td>\n",
       "      <td>positive</td>\n",
       "      <td>0.9412</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Calgary Airport Marriott In-Terminal Hotel</td>\n",
       "      <td>\"Rooms are very modern looking with great king...</td>\n",
       "      <td>5</td>\n",
       "      <td>positive</td>\n",
       "      <td>0.6409</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Calgary Airport Marriott In-Terminal Hotel</td>\n",
       "      <td>\"We stayed just one night at the Marriott and ...</td>\n",
       "      <td>4</td>\n",
       "      <td>positive</td>\n",
       "      <td>0.4957</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Calgary Airport Marriott In-Terminal Hotel</td>\n",
       "      <td>\"The Marriot is across from the old Delta. Bot...</td>\n",
       "      <td>5</td>\n",
       "      <td>positive</td>\n",
       "      <td>0.9735</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 505 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                    hotelName  \\\n",
       "0  Calgary Airport Marriott In-Terminal Hotel   \n",
       "1  Calgary Airport Marriott In-Terminal Hotel   \n",
       "2  Calgary Airport Marriott In-Terminal Hotel   \n",
       "3  Calgary Airport Marriott In-Terminal Hotel   \n",
       "4  Calgary Airport Marriott In-Terminal Hotel   \n",
       "\n",
       "                                        reviewColumn  ratingScore groundTruth  \\\n",
       "0  \"We were excited to stay at the new Marriott i...            5    positive   \n",
       "1  \"We were not able to make our Victoria connect...            5    positive   \n",
       "2  \"Rooms are very modern looking with great king...            5    positive   \n",
       "3  \"We stayed just one night at the Marriott and ...            4    positive   \n",
       "4  \"The Marriot is across from the old Delta. Bot...            5    positive   \n",
       "\n",
       "   Vader_score  front desk  hot tub  downtown calgari  great locat  \\\n",
       "0       0.9830           0        0                 0            0   \n",
       "1       0.9412           0        0                 0            0   \n",
       "2       0.6409           0        0                 0            0   \n",
       "3       0.4957           0        0                 0            0   \n",
       "4       0.9735           0        0                 0            0   \n",
       "\n",
       "   front desk staff     ...      busi center  nearbi restaur  hair dryer  \\\n",
       "0                 0     ...                0               0           0   \n",
       "1                 0     ...                0               0           0   \n",
       "2                 0     ...                0               0           0   \n",
       "3                 0     ...                0               0           0   \n",
       "4                 0     ...                0               0           0   \n",
       "\n",
       "   tran canada highway  waffl maker  room price  riversid inn  larg walk  \\\n",
       "0                    0            0           0             0          1   \n",
       "1                    0            0           0             0          0   \n",
       "2                    0            0           0             0          0   \n",
       "3                    0            0           0             0          0   \n",
       "4                    0            0           0             0          0   \n",
       "\n",
       "   nice clean  numer time  \n",
       "0           0           0  \n",
       "1           0           0  \n",
       "2           0           0  \n",
       "3           0           0  \n",
       "4           0           0  \n",
       "\n",
       "[5 rows x 505 columns]"
      ]
     },
     "execution_count": 136,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "finaldf_phrase.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 137,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[('front desk', 1062),\n",
       " ('hot tub', 446),\n",
       " ('downtown calgari', 318),\n",
       " ('great locat', 302),\n",
       " ('front desk staff', 300),\n",
       " ('nthe room', 259),\n",
       " ('calgari tower', 224),\n",
       " ('room servic', 217),\n",
       " ('great place', 206),\n",
       " ('comfort bed', 200)]"
      ]
     },
     "execution_count": 137,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "topk_phrase[:10]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 138,
   "metadata": {},
   "outputs": [],
   "source": [
    "def topknounphrases(df, label_value='positive',k=50):\n",
    "    hotelDf = df[df['groundTruth'] == label_value]\n",
    "    reviews = hotelDf['reviewColumn'].values\n",
    "    \n",
    "    # Top-k frequent terms\n",
    "    counter = Counter()\n",
    "    for review in reviews:\n",
    "            counter.update(flatten([word\n",
    "                            for word \n",
    "                            in get_terms(chunker.parse(pos_tag(re.findall(r'\\w+', review)))) \n",
    "                            ]))\n",
    "    topk = counter.most_common(k)\n",
    "    \n",
    "    return topk"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 139,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[('front desk', 546),\n",
       " ('hot tub', 325),\n",
       " ('downtown calgari', 273),\n",
       " ('great locat', 255),\n",
       " ('front desk staff', 212),\n",
       " ('calgari tower', 206),\n",
       " ('great place', 192),\n",
       " ('comfort bed', 180),\n",
       " ('nthe room', 175),\n",
       " ('hotel staff', 153),\n",
       " ('great stay', 152),\n",
       " ('room servic', 151),\n",
       " ('breakfast buffet', 141),\n",
       " ('great hotel', 141),\n",
       " ('friendli staff', 139),\n",
       " ('nthe hotel', 137),\n",
       " ('first time', 135),\n",
       " ('good valu', 129),\n",
       " ('nice hotel', 127),\n",
       " ('next time', 127),\n",
       " ('pool area', 118),\n",
       " ('free breakfast', 118),\n",
       " ('good locat', 117),\n",
       " ('short walk', 114),\n",
       " ('hotel room', 110),\n",
       " ('night stay', 109),\n",
       " ('minut walk', 108),\n",
       " ('great servic', 105),\n",
       " ('easi access', 104),\n",
       " ('great view', 99),\n",
       " ('custom servic', 99),\n",
       " ('busi trip', 95),\n",
       " ('next morn', 95),\n",
       " ('good size', 94),\n",
       " ('continent breakfast', 91),\n",
       " ('free park', 91),\n",
       " ('water slide', 91),\n",
       " ('bow river', 90),\n",
       " ('help staff', 89),\n",
       " ('nice touch', 89),\n",
       " ('next day', 86),\n",
       " ('calgari stamped', 86),\n",
       " ('complimentari breakfast', 81),\n",
       " ('calgari airport', 78),\n",
       " ('mani restaur', 78),\n",
       " ('clean room', 77),\n",
       " ('great valu', 76),\n",
       " ('busi travel', 75),\n",
       " ('nthe staff', 75),\n",
       " ('good hotel', 74)]"
      ]
     },
     "execution_count": 139,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "topknounphrases(hotelDf,label_value='positive',k=50)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 140,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[('front desk', 516),\n",
       " ('credit card', 142),\n",
       " ('hot tub', 121),\n",
       " ('next morn', 89),\n",
       " ('front desk staff', 88),\n",
       " ('nthe room', 84),\n",
       " ('next day', 79),\n",
       " ('park lot', 74),\n",
       " ('hotel room', 73),\n",
       " ('first night', 67),\n",
       " ('room servic', 66),\n",
       " ('hot water', 59),\n",
       " ('good locat', 58),\n",
       " ('free breakfast', 55),\n",
       " ('custom servic', 55),\n",
       " ('air condition', 50),\n",
       " ('night stay', 48),\n",
       " ('great locat', 47),\n",
       " ('holiday inn', 47),\n",
       " ('downtown calgari', 45),\n",
       " ('smoke room', 45),\n",
       " ('nthe hotel', 44),\n",
       " ('first time', 42),\n",
       " ('continent breakfast', 42),\n",
       " ('busi travel', 41),\n",
       " ('non smoke room', 39),\n",
       " ('good thing', 38),\n",
       " ('queen bed', 37),\n",
       " ('next time', 37),\n",
       " ('hotel staff', 37),\n",
       " ('staff member', 36),\n",
       " ('pool area', 36),\n",
       " ('key card', 34),\n",
       " ('second night', 33),\n",
       " ('desk clerk', 32),\n",
       " ('minut walk', 32),\n",
       " ('long time', 32),\n",
       " ('second floor', 31),\n",
       " ('nice hotel', 31),\n",
       " ('breakfast room', 31),\n",
       " ('ice machin', 31),\n",
       " ('good valu', 31),\n",
       " ('busi trip', 31),\n",
       " ('complimentari breakfast', 31),\n",
       " ('differ room', 31),\n",
       " ('clean staff', 31),\n",
       " ('clean room', 29),\n",
       " ('coffe maker', 28),\n",
       " ('room key', 28),\n",
       " ('phone call', 27)]"
      ]
     },
     "execution_count": 140,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "topknounphrases(hotelDf,label_value='negative',k=50)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Insights about Top Ranked Positive & Negative review Noun Phrases:"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Positive: In positive reviews we have top phrases as \"downtown calgari\"/\"great location\" etc. which shows the general importance of location that is giving positive reviews. Similarly in negative reviews \"credit card\"/\"parking lot\" has frequent mentions.\n",
    "\n",
    "Interestingly, both in positive and negative cases we can find that we have common phrases like **front desk/hot tub/room service** etc. Since Top Key/Noun phrases indicates the most frequent terms that the customers are mentioning in the reviews, In our case we can say that these the things that the customers are more concerned. Like explained previously they might be positive and negative mentions in both the cases respectively.\n",
    "\n",
    "In our previous case, while analyzing the individual words **staff** was identified most frequent in both positive and negative reviews. But from the phrasal analysis we can further say that it is about **front desk staff** that we have more mentions.\n",
    "\n",
    "\n",
    "All the above phrases/words impotance can be better explored by calculating Mutual information and PMI."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Q3. Mutual Information"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 141,
   "metadata": {},
   "outputs": [],
   "source": [
    "# get Top K mutual information terms from the dataframe\n",
    "def getMI(topk, df, label_column='groundTruth'):\n",
    "    miScore = []\n",
    "    for word in topk:\n",
    "        miScore.append([word[0]]+[metrics.mutual_info_score(df[label_column], df[word[0]])])\n",
    "    miScoredf = pandas.DataFrame(miScore).sort_values(1,ascending=0)\n",
    "    miScoredf.columns = ['Word','MI Score']\n",
    "    return miScoredf"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 142,
   "metadata": {},
   "outputs": [],
   "source": [
    "miScoredf = getMI(topk, finaldf)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 147,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Word</th>\n",
       "      <th>MI Score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>great</td>\n",
       "      <td>0.028662</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>282</th>\n",
       "      <td>dirty</td>\n",
       "      <td>0.026815</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>114</th>\n",
       "      <td>told</td>\n",
       "      <td>0.025522</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>125</th>\n",
       "      <td>said</td>\n",
       "      <td>0.018058</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>42</th>\n",
       "      <td>excellent</td>\n",
       "      <td>0.016005</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>185</th>\n",
       "      <td>called</td>\n",
       "      <td>0.013587</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>408</th>\n",
       "      <td>poor</td>\n",
       "      <td>0.013266</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>135</th>\n",
       "      <td>old</td>\n",
       "      <td>0.013041</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>130</th>\n",
       "      <td>asked</td>\n",
       "      <td>0.012880</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>room</td>\n",
       "      <td>0.011559</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>night</td>\n",
       "      <td>0.011017</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>104</th>\n",
       "      <td>never</td>\n",
       "      <td>0.010999</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>comfortable</td>\n",
       "      <td>0.010626</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>186</th>\n",
       "      <td>bad</td>\n",
       "      <td>0.010485</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>119</th>\n",
       "      <td>enjoyed</td>\n",
       "      <td>0.010300</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>calgary</td>\n",
       "      <td>0.010025</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>242</th>\n",
       "      <td>card</td>\n",
       "      <td>0.009378</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>359</th>\n",
       "      <td>someone</td>\n",
       "      <td>0.009328</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>139</th>\n",
       "      <td>another</td>\n",
       "      <td>0.008814</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>80</th>\n",
       "      <td>door</td>\n",
       "      <td>0.008573</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>95</th>\n",
       "      <td>spacious</td>\n",
       "      <td>0.008543</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>56</th>\n",
       "      <td>restaurants</td>\n",
       "      <td>0.008418</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>466</th>\n",
       "      <td>smoking</td>\n",
       "      <td>0.008332</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>167</th>\n",
       "      <td>left</td>\n",
       "      <td>0.008231</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>81</th>\n",
       "      <td>booked</td>\n",
       "      <td>0.008226</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>30</th>\n",
       "      <td>helpful</td>\n",
       "      <td>0.008219</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>393</th>\n",
       "      <td>credit</td>\n",
       "      <td>0.008178</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>250</th>\n",
       "      <td>pay</td>\n",
       "      <td>0.008170</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>220</th>\n",
       "      <td>highly</td>\n",
       "      <td>0.008114</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>169</th>\n",
       "      <td>nothing</td>\n",
       "      <td>0.008108</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>142</th>\n",
       "      <td>wonderful</td>\n",
       "      <td>0.008075</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>staff</td>\n",
       "      <td>0.007981</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>friendly</td>\n",
       "      <td>0.007884</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>get</td>\n",
       "      <td>0.007801</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>36</th>\n",
       "      <td>could</td>\n",
       "      <td>0.007647</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>233</th>\n",
       "      <td>air</td>\n",
       "      <td>0.007467</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>429</th>\n",
       "      <td>toilet</td>\n",
       "      <td>0.007415</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>one</td>\n",
       "      <td>0.007391</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>94</th>\n",
       "      <td>people</td>\n",
       "      <td>0.007351</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>457</th>\n",
       "      <td>sheets</td>\n",
       "      <td>0.007286</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>155</th>\n",
       "      <td>perfect</td>\n",
       "      <td>0.007248</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>320</th>\n",
       "      <td>paid</td>\n",
       "      <td>0.006985</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>68</th>\n",
       "      <td>definitely</td>\n",
       "      <td>0.006967</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>115</th>\n",
       "      <td>however</td>\n",
       "      <td>0.006957</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>217</th>\n",
       "      <td>manager</td>\n",
       "      <td>0.006865</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>quot</td>\n",
       "      <td>0.006723</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>128</th>\n",
       "      <td>amazing</td>\n",
       "      <td>0.006647</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>46</th>\n",
       "      <td>even</td>\n",
       "      <td>0.006536</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>306</th>\n",
       "      <td>phone</td>\n",
       "      <td>0.006467</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>424</th>\n",
       "      <td>tried</td>\n",
       "      <td>0.006398</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            Word  MI Score\n",
       "4          great  0.028662\n",
       "282        dirty  0.026815\n",
       "114         told  0.025522\n",
       "125         said  0.018058\n",
       "42     excellent  0.016005\n",
       "185       called  0.013587\n",
       "408         poor  0.013266\n",
       "135          old  0.013041\n",
       "130        asked  0.012880\n",
       "1           room  0.011559\n",
       "16         night  0.011017\n",
       "104        never  0.010999\n",
       "17   comfortable  0.010626\n",
       "186          bad  0.010485\n",
       "119      enjoyed  0.010300\n",
       "7        calgary  0.010025\n",
       "242         card  0.009378\n",
       "359      someone  0.009328\n",
       "139      another  0.008814\n",
       "80          door  0.008573\n",
       "95      spacious  0.008543\n",
       "56   restaurants  0.008418\n",
       "466      smoking  0.008332\n",
       "167         left  0.008231\n",
       "81        booked  0.008226\n",
       "30       helpful  0.008219\n",
       "393       credit  0.008178\n",
       "250          pay  0.008170\n",
       "220       highly  0.008114\n",
       "169      nothing  0.008108\n",
       "142    wonderful  0.008075\n",
       "2          staff  0.007981\n",
       "15      friendly  0.007884\n",
       "22           get  0.007801\n",
       "36         could  0.007647\n",
       "233          air  0.007467\n",
       "429       toilet  0.007415\n",
       "13           one  0.007391\n",
       "94        people  0.007351\n",
       "457       sheets  0.007286\n",
       "155      perfect  0.007248\n",
       "320         paid  0.006985\n",
       "68    definitely  0.006967\n",
       "115      however  0.006957\n",
       "217      manager  0.006865\n",
       "29          quot  0.006723\n",
       "128      amazing  0.006647\n",
       "46          even  0.006536\n",
       "306        phone  0.006467\n",
       "424        tried  0.006398"
      ]
     },
     "execution_count": 147,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "miScoredf.head(50)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Using mutual information (MI) with ground truth sentiment we can see that the top-50 most sentiment-bearing words include terms like great/dirty/excellent etc. These words are self explainatory because they cleary weighs for either a positive or negative sentiment. (Considering non negation cases)\n",
    "\n",
    "Interestingly, we can also find words like **told/said/called** etc which looks like stop words. But the reason might be that these words might be predictive for negative cases because the words shows complaining behaviour. We can confirm this by checking for their PMI in the negative cases."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [],
   "source": [
    "miScoredf_phrase = getMI(topk_phrase, finaldf_phrase)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 125,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Word</th>\n",
       "      <th>MI Score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>credit card</td>\n",
       "      <td>0.007879</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>front desk</td>\n",
       "      <td>0.005631</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>92</th>\n",
       "      <td>air condition</td>\n",
       "      <td>0.002621</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>161</th>\n",
       "      <td>non smoke room</td>\n",
       "      <td>0.002467</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>great stay</td>\n",
       "      <td>0.002315</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>198</th>\n",
       "      <td>key card</td>\n",
       "      <td>0.002156</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>121</th>\n",
       "      <td>smoke room</td>\n",
       "      <td>0.002130</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>great place</td>\n",
       "      <td>0.001885</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>calgari tower</td>\n",
       "      <td>0.001742</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>park lot</td>\n",
       "      <td>0.001693</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>152</th>\n",
       "      <td>second night</td>\n",
       "      <td>0.001690</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>49</th>\n",
       "      <td>hot water</td>\n",
       "      <td>0.001658</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>242</th>\n",
       "      <td>plu side</td>\n",
       "      <td>0.001527</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>great hotel</td>\n",
       "      <td>0.001457</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>next morn</td>\n",
       "      <td>0.001344</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>155</th>\n",
       "      <td>differ room</td>\n",
       "      <td>0.001307</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31</th>\n",
       "      <td>first night</td>\n",
       "      <td>0.001294</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>next day</td>\n",
       "      <td>0.001288</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>456</th>\n",
       "      <td>bed sheet</td>\n",
       "      <td>0.001272</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>156</th>\n",
       "      <td>ice machin</td>\n",
       "      <td>0.001243</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>95</th>\n",
       "      <td>good thing</td>\n",
       "      <td>0.001239</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>comfort bed</td>\n",
       "      <td>0.001238</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>44</th>\n",
       "      <td>great view</td>\n",
       "      <td>0.001176</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>454</th>\n",
       "      <td>secur deposit</td>\n",
       "      <td>0.001174</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>139</th>\n",
       "      <td>desk clerk</td>\n",
       "      <td>0.001111</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>52</th>\n",
       "      <td>bow river</td>\n",
       "      <td>0.001099</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>268</th>\n",
       "      <td>bathroom floor</td>\n",
       "      <td>0.001095</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>485</th>\n",
       "      <td>bed bug</td>\n",
       "      <td>0.001035</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37</th>\n",
       "      <td>great servic</td>\n",
       "      <td>0.001020</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>183</th>\n",
       "      <td>phone call</td>\n",
       "      <td>0.001007</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>61</th>\n",
       "      <td>mani restaur</td>\n",
       "      <td>0.000994</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>119</th>\n",
       "      <td>long time</td>\n",
       "      <td>0.000943</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>63</th>\n",
       "      <td>great valu</td>\n",
       "      <td>0.000936</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>downtown calgari</td>\n",
       "      <td>0.000924</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>345</th>\n",
       "      <td>front desk person</td>\n",
       "      <td>0.000848</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>148</th>\n",
       "      <td>wonder stay</td>\n",
       "      <td>0.000823</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>208</th>\n",
       "      <td>beauti hotel</td>\n",
       "      <td>0.000818</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>226</th>\n",
       "      <td>stephen avenu</td>\n",
       "      <td>0.000795</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>210</th>\n",
       "      <td>first room</td>\n",
       "      <td>0.000778</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>97</th>\n",
       "      <td>top notch</td>\n",
       "      <td>0.000774</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>325</th>\n",
       "      <td>cell phone</td>\n",
       "      <td>0.000769</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>356</th>\n",
       "      <td>second room</td>\n",
       "      <td>0.000769</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>53</th>\n",
       "      <td>holiday inn</td>\n",
       "      <td>0.000766</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>464</th>\n",
       "      <td>cigarett smoke</td>\n",
       "      <td>0.000758</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>125</th>\n",
       "      <td>spaciou room</td>\n",
       "      <td>0.000756</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>82</th>\n",
       "      <td>club loung</td>\n",
       "      <td>0.000735</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>great locat</td>\n",
       "      <td>0.000727</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>247</th>\n",
       "      <td>excel valu</td>\n",
       "      <td>0.000724</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>328</th>\n",
       "      <td>small room</td>\n",
       "      <td>0.000702</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>383</th>\n",
       "      <td>next room</td>\n",
       "      <td>0.000691</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                  Word  MI Score\n",
       "17         credit card  0.007879\n",
       "0           front desk  0.005631\n",
       "92       air condition  0.002621\n",
       "161     non smoke room  0.002467\n",
       "25          great stay  0.002315\n",
       "198           key card  0.002156\n",
       "121         smoke room  0.002130\n",
       "8          great place  0.001885\n",
       "6        calgari tower  0.001742\n",
       "29            park lot  0.001693\n",
       "152       second night  0.001690\n",
       "49           hot water  0.001658\n",
       "242           plu side  0.001527\n",
       "28         great hotel  0.001457\n",
       "11           next morn  0.001344\n",
       "155        differ room  0.001307\n",
       "31         first night  0.001294\n",
       "18            next day  0.001288\n",
       "456          bed sheet  0.001272\n",
       "156         ice machin  0.001243\n",
       "95          good thing  0.001239\n",
       "9          comfort bed  0.001238\n",
       "44          great view  0.001176\n",
       "454      secur deposit  0.001174\n",
       "139         desk clerk  0.001111\n",
       "52           bow river  0.001099\n",
       "268     bathroom floor  0.001095\n",
       "485            bed bug  0.001035\n",
       "37        great servic  0.001020\n",
       "183         phone call  0.001007\n",
       "61        mani restaur  0.000994\n",
       "119          long time  0.000943\n",
       "63          great valu  0.000936\n",
       "2     downtown calgari  0.000924\n",
       "345  front desk person  0.000848\n",
       "148        wonder stay  0.000823\n",
       "208       beauti hotel  0.000818\n",
       "226      stephen avenu  0.000795\n",
       "210         first room  0.000778\n",
       "97           top notch  0.000774\n",
       "325         cell phone  0.000769\n",
       "356        second room  0.000769\n",
       "53         holiday inn  0.000766\n",
       "464     cigarett smoke  0.000758\n",
       "125       spaciou room  0.000756\n",
       "82          club loung  0.000735\n",
       "3          great locat  0.000727\n",
       "247         excel valu  0.000724\n",
       "328         small room  0.000702\n",
       "383          next room  0.000691"
      ]
     },
     "execution_count": 125,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "miScoredf_phrase.head(50)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can find the phrases like \"credit card\"/ \"front desk\" / \"non smoke room\" with high mutual information - impling that they contribute more for predicting the correct class (Also, these words are self explainatory Eg: Credit card - Customers complaining about it (Negative)).\n",
    "\n",
    "Also, we can see that MI values for phrases are lower than individual words this might be because MI considers frequency while making calculations and since similar phrases frequencies will be generally lower they corresponds to lower MI"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Q4. Pointwise Mutual Information"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Simple example of getting pairwise mutual information of a term\n",
    "def pmiCal(df, x):\n",
    "    pmilist=[]\n",
    "    for i in ['positive','negative']:\n",
    "        for j in [0,1]:\n",
    "            px = sum(df['groundTruth']==i)/len(df)\n",
    "            py = sum(df[x]==j)/len(df)\n",
    "            pxy = len(df[(df['groundTruth']==i) & (df[x]==j)])/len(df)\n",
    "            if px or py == 0:\n",
    "                pmi = math.log((pxy+0.0001)/(px*py+0.0001))\n",
    "            \n",
    "            elif pxy==0:#Log 0 cannot happen\n",
    "                pmi = math.log((pxy+0.0001)/(px*py+0.0001))\n",
    "            \n",
    "            else:\n",
    "                pmi = math.log(pxy/(px*py))\n",
    "            pmilist.append([i]+[j]+[px]+[py]+[pxy]+[pmi])\n",
    "    pmidf = pandas.DataFrame(pmilist)\n",
    "    pmidf.columns = ['x','y','px','py','pxy','pmi']\n",
    "    return pmidf"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [],
   "source": [
    "def pmiIndivCal(df,x,gt, label_column='groundTruth'):\n",
    "    px = sum(df[label_column]==gt)/len(df)\n",
    "    py = sum(df[x]==1)/len(df)\n",
    "    pxy = len(df[(df[label_column]==gt) & (df[x]==1)])/len(df)\n",
    "    \n",
    "    if px*py == 0:\n",
    "        pmi = math.log((pxy+0.0001)/(px*py+0.0001))\n",
    "        \n",
    "        return pmi\n",
    "    elif pxy==0:#Log 0 cannot happen\n",
    "        pmi = math.log((pxy+0.0001)/(px*py+0.0001))\n",
    "        return pmi\n",
    "    else:\n",
    "        pmi = math.log(pxy/(px*py))\n",
    "    return pmi"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Compute PMI for all terms and all possible labels\n",
    "def pmiForAllCal(df, label_column='groundTruth', topk=topk):\n",
    "    #Try calculate all the pmi for top k and store them into one pmidf dataframe\n",
    "    pmilist = []\n",
    "    pmiposlist = []\n",
    "    pmineglist = []\n",
    "    for word in tqdm(topk):\n",
    "        pmilist.append([word[0]]+[pmiCal(df,word[0])])\n",
    "        pmiposlist.append([word[0]]+[pmiIndivCal(df,word[0],'positive',label_column)])\n",
    "        pmineglist.append([word[0]]+[pmiIndivCal(df,word[0],'negative',label_column)])\n",
    "    pmidf = pandas.DataFrame(pmilist)\n",
    "    pmiposlist = pandas.DataFrame(pmiposlist)\n",
    "    pmineglist = pandas.DataFrame(pmineglist)\n",
    "    pmiposlist.columns = ['word','pmi']\n",
    "    pmineglist.columns = ['word','pmi']\n",
    "    pmidf.columns = ['word','pmi']\n",
    "    return pmiposlist, pmineglist, pmidf"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|████████████████████████████████████████████████████████████████████████████████| 500/500 [00:32<00:00, 15.21it/s]\n"
     ]
    }
   ],
   "source": [
    "pmiposlist, pmineglist, pmidf = pmiForAllCal(finaldf)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 144,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>word</th>\n",
       "      <th>pmi</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>220</th>\n",
       "      <td>highly</td>\n",
       "      <td>0.261977</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>119</th>\n",
       "      <td>enjoyed</td>\n",
       "      <td>0.244841</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>380</th>\n",
       "      <td>delicious</td>\n",
       "      <td>0.241703</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>142</th>\n",
       "      <td>wonderful</td>\n",
       "      <td>0.241383</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>155</th>\n",
       "      <td>perfect</td>\n",
       "      <td>0.235798</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>263</th>\n",
       "      <td>awesome</td>\n",
       "      <td>0.233271</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>197</th>\n",
       "      <td>loved</td>\n",
       "      <td>0.233053</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>202</th>\n",
       "      <td>beautiful</td>\n",
       "      <td>0.230314</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>277</th>\n",
       "      <td>river</td>\n",
       "      <td>0.226056</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>128</th>\n",
       "      <td>amazing</td>\n",
       "      <td>0.221182</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>42</th>\n",
       "      <td>excellent</td>\n",
       "      <td>0.215121</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>95</th>\n",
       "      <td>spacious</td>\n",
       "      <td>0.208493</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>479</th>\n",
       "      <td>thanks</td>\n",
       "      <td>0.205946</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>274</th>\n",
       "      <td>love</td>\n",
       "      <td>0.204041</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>411</th>\n",
       "      <td>attentive</td>\n",
       "      <td>0.204005</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>289</th>\n",
       "      <td>accommodating</td>\n",
       "      <td>0.202011</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>353</th>\n",
       "      <td>appointed</td>\n",
       "      <td>0.201755</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>223</th>\n",
       "      <td>fantastic</td>\n",
       "      <td>0.199431</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>436</th>\n",
       "      <td>shops</td>\n",
       "      <td>0.193386</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>395</th>\n",
       "      <td>welcome</td>\n",
       "      <td>0.191681</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>386</th>\n",
       "      <td>welcoming</td>\n",
       "      <td>0.187104</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>56</th>\n",
       "      <td>restaurants</td>\n",
       "      <td>0.182642</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>471</th>\n",
       "      <td>visiting</td>\n",
       "      <td>0.181167</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>182</th>\n",
       "      <td>lovely</td>\n",
       "      <td>0.180358</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>496</th>\n",
       "      <td>beyond</td>\n",
       "      <td>0.179740</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>68</th>\n",
       "      <td>definitely</td>\n",
       "      <td>0.179390</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>308</th>\n",
       "      <td>efficient</td>\n",
       "      <td>0.179099</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>419</th>\n",
       "      <td>courteous</td>\n",
       "      <td>0.179021</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>107</th>\n",
       "      <td>modern</td>\n",
       "      <td>0.178448</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>480</th>\n",
       "      <td>concierge</td>\n",
       "      <td>0.177038</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>312</th>\n",
       "      <td>thank</td>\n",
       "      <td>0.172399</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>290</th>\n",
       "      <td>marriott</td>\n",
       "      <td>0.171945</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>295</th>\n",
       "      <td>club</td>\n",
       "      <td>0.171380</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>399</th>\n",
       "      <td>mall</td>\n",
       "      <td>0.169649</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>106</th>\n",
       "      <td>lounge</td>\n",
       "      <td>0.168109</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>336</th>\n",
       "      <td>professional</td>\n",
       "      <td>0.166265</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>great</td>\n",
       "      <td>0.165963</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>78</th>\n",
       "      <td>easy</td>\n",
       "      <td>0.163600</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>176</th>\n",
       "      <td>shopping</td>\n",
       "      <td>0.163086</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>239</th>\n",
       "      <td>tower</td>\n",
       "      <td>0.161781</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>381</th>\n",
       "      <td>starbucks</td>\n",
       "      <td>0.160986</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>441</th>\n",
       "      <td>fun</td>\n",
       "      <td>0.158073</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>490</th>\n",
       "      <td>variety</td>\n",
       "      <td>0.157017</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>141</th>\n",
       "      <td>distance</td>\n",
       "      <td>0.150257</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>343</th>\n",
       "      <td>sheraton</td>\n",
       "      <td>0.147097</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>63</th>\n",
       "      <td>always</td>\n",
       "      <td>0.145737</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>30</th>\n",
       "      <td>helpful</td>\n",
       "      <td>0.144188</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>71</th>\n",
       "      <td>quiet</td>\n",
       "      <td>0.142076</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>367</th>\n",
       "      <td>enjoy</td>\n",
       "      <td>0.141294</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>365</th>\n",
       "      <td>facilities</td>\n",
       "      <td>0.136774</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              word       pmi\n",
       "220         highly  0.261977\n",
       "119        enjoyed  0.244841\n",
       "380      delicious  0.241703\n",
       "142      wonderful  0.241383\n",
       "155        perfect  0.235798\n",
       "263        awesome  0.233271\n",
       "197          loved  0.233053\n",
       "202      beautiful  0.230314\n",
       "277          river  0.226056\n",
       "128        amazing  0.221182\n",
       "42       excellent  0.215121\n",
       "95        spacious  0.208493\n",
       "479         thanks  0.205946\n",
       "274           love  0.204041\n",
       "411      attentive  0.204005\n",
       "289  accommodating  0.202011\n",
       "353      appointed  0.201755\n",
       "223      fantastic  0.199431\n",
       "436          shops  0.193386\n",
       "395        welcome  0.191681\n",
       "386      welcoming  0.187104\n",
       "56     restaurants  0.182642\n",
       "471       visiting  0.181167\n",
       "182         lovely  0.180358\n",
       "496         beyond  0.179740\n",
       "68      definitely  0.179390\n",
       "308      efficient  0.179099\n",
       "419      courteous  0.179021\n",
       "107         modern  0.178448\n",
       "480      concierge  0.177038\n",
       "312          thank  0.172399\n",
       "290       marriott  0.171945\n",
       "295           club  0.171380\n",
       "399           mall  0.169649\n",
       "106         lounge  0.168109\n",
       "336   professional  0.166265\n",
       "4            great  0.165963\n",
       "78            easy  0.163600\n",
       "176       shopping  0.163086\n",
       "239          tower  0.161781\n",
       "381      starbucks  0.160986\n",
       "441            fun  0.158073\n",
       "490        variety  0.157017\n",
       "141       distance  0.150257\n",
       "343       sheraton  0.147097\n",
       "63          always  0.145737\n",
       "30         helpful  0.144188\n",
       "71           quiet  0.142076\n",
       "367          enjoy  0.141294\n",
       "365     facilities  0.136774"
      ]
     },
     "execution_count": 144,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Sorted top pmi words for positive reviews\n",
    "pmiposlist.sort_values('pmi',ascending=0).head(50)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "for the top PMI words in the positive reviews we can find words like \"spacious\"/ \"river\" / \"delicous\" etc which mentions about the location, ambience or food which are accounting for the positive reviews."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 149,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>word</th>\n",
       "      <th>pmi</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>282</th>\n",
       "      <td>dirty</td>\n",
       "      <td>1.264574</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>466</th>\n",
       "      <td>smoking</td>\n",
       "      <td>1.109784</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>408</th>\n",
       "      <td>poor</td>\n",
       "      <td>1.105982</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>114</th>\n",
       "      <td>told</td>\n",
       "      <td>1.033153</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>457</th>\n",
       "      <td>sheets</td>\n",
       "      <td>1.005240</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>393</th>\n",
       "      <td>credit</td>\n",
       "      <td>1.004029</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>429</th>\n",
       "      <td>toilet</td>\n",
       "      <td>0.975498</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>359</th>\n",
       "      <td>someone</td>\n",
       "      <td>0.942464</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>125</th>\n",
       "      <td>said</td>\n",
       "      <td>0.938461</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>242</th>\n",
       "      <td>card</td>\n",
       "      <td>0.914747</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>185</th>\n",
       "      <td>called</td>\n",
       "      <td>0.913292</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>438</th>\n",
       "      <td>non</td>\n",
       "      <td>0.894244</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>424</th>\n",
       "      <td>tried</td>\n",
       "      <td>0.893437</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>383</th>\n",
       "      <td>loud</td>\n",
       "      <td>0.847706</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>306</th>\n",
       "      <td>phone</td>\n",
       "      <td>0.830364</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>320</th>\n",
       "      <td>paid</td>\n",
       "      <td>0.823717</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>135</th>\n",
       "      <td>old</td>\n",
       "      <td>0.819227</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>250</th>\n",
       "      <td>pay</td>\n",
       "      <td>0.813260</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>186</th>\n",
       "      <td>bad</td>\n",
       "      <td>0.813209</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>420</th>\n",
       "      <td>wall</td>\n",
       "      <td>0.812705</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>130</th>\n",
       "      <td>asked</td>\n",
       "      <td>0.805969</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>402</th>\n",
       "      <td>dated</td>\n",
       "      <td>0.803135</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>394</th>\n",
       "      <td>later</td>\n",
       "      <td>0.794537</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>217</th>\n",
       "      <td>manager</td>\n",
       "      <td>0.783514</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>233</th>\n",
       "      <td>air</td>\n",
       "      <td>0.776810</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>341</th>\n",
       "      <td>average</td>\n",
       "      <td>0.766475</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>459</th>\n",
       "      <td>basic</td>\n",
       "      <td>0.745264</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>281</th>\n",
       "      <td>put</td>\n",
       "      <td>0.738930</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>475</th>\n",
       "      <td>cleaning</td>\n",
       "      <td>0.737171</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>461</th>\n",
       "      <td>either</td>\n",
       "      <td>0.735854</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>450</th>\n",
       "      <td>key</td>\n",
       "      <td>0.732717</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>232</th>\n",
       "      <td>towels</td>\n",
       "      <td>0.725926</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>316</th>\n",
       "      <td>noisy</td>\n",
       "      <td>0.724118</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>167</th>\n",
       "      <td>left</td>\n",
       "      <td>0.719226</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>430</th>\n",
       "      <td>window</td>\n",
       "      <td>0.713041</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>169</th>\n",
       "      <td>nothing</td>\n",
       "      <td>0.708741</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>104</th>\n",
       "      <td>never</td>\n",
       "      <td>0.707050</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>139</th>\n",
       "      <td>another</td>\n",
       "      <td>0.702080</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>302</th>\n",
       "      <td>seemed</td>\n",
       "      <td>0.701977</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>291</th>\n",
       "      <td>motel</td>\n",
       "      <td>0.700925</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>444</th>\n",
       "      <td>change</td>\n",
       "      <td>0.688036</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>493</th>\n",
       "      <td>elevators</td>\n",
       "      <td>0.673408</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>361</th>\n",
       "      <td>reservation</td>\n",
       "      <td>0.672535</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>434</th>\n",
       "      <td>least</td>\n",
       "      <td>0.668768</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>476</th>\n",
       "      <td>doors</td>\n",
       "      <td>0.655464</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>294</th>\n",
       "      <td>working</td>\n",
       "      <td>0.650796</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>451</th>\n",
       "      <td>run</td>\n",
       "      <td>0.649574</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>210</th>\n",
       "      <td>call</td>\n",
       "      <td>0.648520</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>261</th>\n",
       "      <td>charge</td>\n",
       "      <td>0.642100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>322</th>\n",
       "      <td>person</td>\n",
       "      <td>0.628648</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            word       pmi\n",
       "282        dirty  1.264574\n",
       "466      smoking  1.109784\n",
       "408         poor  1.105982\n",
       "114         told  1.033153\n",
       "457       sheets  1.005240\n",
       "393       credit  1.004029\n",
       "429       toilet  0.975498\n",
       "359      someone  0.942464\n",
       "125         said  0.938461\n",
       "242         card  0.914747\n",
       "185       called  0.913292\n",
       "438          non  0.894244\n",
       "424        tried  0.893437\n",
       "383         loud  0.847706\n",
       "306        phone  0.830364\n",
       "320         paid  0.823717\n",
       "135          old  0.819227\n",
       "250          pay  0.813260\n",
       "186          bad  0.813209\n",
       "420         wall  0.812705\n",
       "130        asked  0.805969\n",
       "402        dated  0.803135\n",
       "394        later  0.794537\n",
       "217      manager  0.783514\n",
       "233          air  0.776810\n",
       "341      average  0.766475\n",
       "459        basic  0.745264\n",
       "281          put  0.738930\n",
       "475     cleaning  0.737171\n",
       "461       either  0.735854\n",
       "450          key  0.732717\n",
       "232       towels  0.725926\n",
       "316        noisy  0.724118\n",
       "167         left  0.719226\n",
       "430       window  0.713041\n",
       "169      nothing  0.708741\n",
       "104        never  0.707050\n",
       "139      another  0.702080\n",
       "302       seemed  0.701977\n",
       "291        motel  0.700925\n",
       "444       change  0.688036\n",
       "493    elevators  0.673408\n",
       "361  reservation  0.672535\n",
       "434        least  0.668768\n",
       "476        doors  0.655464\n",
       "294      working  0.650796\n",
       "451          run  0.649574\n",
       "210         call  0.648520\n",
       "261       charge  0.642100\n",
       "322       person  0.628648"
      ]
     },
     "execution_count": 149,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Sorted top pmi words for negative reviews\n",
    "pmineglist.sort_values('pmi',ascending=0).head(50)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As hypothesised previously words like said/told/called etc which have **high MI also have high PMI in the negative cases**. This confirms about the complaining behaviour and they should not be considered as stop words."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 150,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|████████████████████████████████████████████████████████████████████████████████| 500/500 [01:00<00:00,  8.16it/s]\n"
     ]
    }
   ],
   "source": [
    "pmiposlist_phrase, pmineglist_phrase, pmidf_phrase = pmiForAllCal(finaldf_phrase, topk=topk_phrase)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 151,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>word</th>\n",
       "      <th>pmi</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>397</th>\n",
       "      <td>super clean</td>\n",
       "      <td>0.295942</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>415</th>\n",
       "      <td>great select</td>\n",
       "      <td>0.295942</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>226</th>\n",
       "      <td>stephen avenu</td>\n",
       "      <td>0.295942</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>468</th>\n",
       "      <td>execut loung</td>\n",
       "      <td>0.295942</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>461</th>\n",
       "      <td>amaz staff</td>\n",
       "      <td>0.295942</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>247</th>\n",
       "      <td>excel valu</td>\n",
       "      <td>0.295942</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>208</th>\n",
       "      <td>beauti hotel</td>\n",
       "      <td>0.295942</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>354</th>\n",
       "      <td>excel place</td>\n",
       "      <td>0.295942</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>479</th>\n",
       "      <td>wonder time</td>\n",
       "      <td>0.295942</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>400</th>\n",
       "      <td>coffe shop</td>\n",
       "      <td>0.295942</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>277</th>\n",
       "      <td>wonder experi</td>\n",
       "      <td>0.295942</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>445</th>\n",
       "      <td>excel choic</td>\n",
       "      <td>0.295942</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>439</th>\n",
       "      <td>entir staff</td>\n",
       "      <td>0.295942</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>463</th>\n",
       "      <td>fantast stay</td>\n",
       "      <td>0.295942</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>433</th>\n",
       "      <td>hotel le germain</td>\n",
       "      <td>0.295942</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>498</th>\n",
       "      <td>nice clean</td>\n",
       "      <td>0.295942</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>148</th>\n",
       "      <td>wonder stay</td>\n",
       "      <td>0.274889</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>228</th>\n",
       "      <td>excel staff</td>\n",
       "      <td>0.265171</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>237</th>\n",
       "      <td>great price</td>\n",
       "      <td>0.265171</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>239</th>\n",
       "      <td>great choic</td>\n",
       "      <td>0.264194</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>great stay</td>\n",
       "      <td>0.263153</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>255</th>\n",
       "      <td>princ island park</td>\n",
       "      <td>0.262041</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>244</th>\n",
       "      <td>extra mile</td>\n",
       "      <td>0.262041</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>269</th>\n",
       "      <td>enjoy stay</td>\n",
       "      <td>0.260851</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>125</th>\n",
       "      <td>spaciou room</td>\n",
       "      <td>0.258202</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>348</th>\n",
       "      <td>wonder hotel</td>\n",
       "      <td>0.253383</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>347</th>\n",
       "      <td>ad bonu</td>\n",
       "      <td>0.253383</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>367</th>\n",
       "      <td>warm welcom</td>\n",
       "      <td>0.251491</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>158</th>\n",
       "      <td>help room</td>\n",
       "      <td>0.250480</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>97</th>\n",
       "      <td>top notch</td>\n",
       "      <td>0.247152</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>379</th>\n",
       "      <td>hot item</td>\n",
       "      <td>0.247152</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>61</th>\n",
       "      <td>mani restaur</td>\n",
       "      <td>0.245932</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>82</th>\n",
       "      <td>club loung</td>\n",
       "      <td>0.245511</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>396</th>\n",
       "      <td>lobbi bar</td>\n",
       "      <td>0.244649</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>188</th>\n",
       "      <td>excel hotel</td>\n",
       "      <td>0.244649</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>44</th>\n",
       "      <td>great view</td>\n",
       "      <td>0.244117</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>63</th>\n",
       "      <td>great valu</td>\n",
       "      <td>0.243983</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>52</th>\n",
       "      <td>bow river</td>\n",
       "      <td>0.241875</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>200</th>\n",
       "      <td>nice stay</td>\n",
       "      <td>0.240372</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>489</th>\n",
       "      <td>fit facil</td>\n",
       "      <td>0.238784</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>478</th>\n",
       "      <td>restaur bar</td>\n",
       "      <td>0.238784</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>470</th>\n",
       "      <td>first class</td>\n",
       "      <td>0.238784</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>483</th>\n",
       "      <td>free upgrad</td>\n",
       "      <td>0.238784</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>211</th>\n",
       "      <td>great food</td>\n",
       "      <td>0.237102</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>220</th>\n",
       "      <td>perfect locat</td>\n",
       "      <td>0.235318</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>215</th>\n",
       "      <td>kensington riversid inn</td>\n",
       "      <td>0.235318</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>213</th>\n",
       "      <td>carriag hous inn</td>\n",
       "      <td>0.235318</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>497</th>\n",
       "      <td>larg walk</td>\n",
       "      <td>0.235318</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>435</th>\n",
       "      <td>chef tabl</td>\n",
       "      <td>0.235318</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>473</th>\n",
       "      <td>club level</td>\n",
       "      <td>0.231404</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                        word       pmi\n",
       "397              super clean  0.295942\n",
       "415             great select  0.295942\n",
       "226            stephen avenu  0.295942\n",
       "468             execut loung  0.295942\n",
       "461               amaz staff  0.295942\n",
       "247               excel valu  0.295942\n",
       "208             beauti hotel  0.295942\n",
       "354              excel place  0.295942\n",
       "479              wonder time  0.295942\n",
       "400               coffe shop  0.295942\n",
       "277            wonder experi  0.295942\n",
       "445              excel choic  0.295942\n",
       "439              entir staff  0.295942\n",
       "463             fantast stay  0.295942\n",
       "433         hotel le germain  0.295942\n",
       "498               nice clean  0.295942\n",
       "148              wonder stay  0.274889\n",
       "228              excel staff  0.265171\n",
       "237              great price  0.265171\n",
       "239              great choic  0.264194\n",
       "25                great stay  0.263153\n",
       "255        princ island park  0.262041\n",
       "244               extra mile  0.262041\n",
       "269               enjoy stay  0.260851\n",
       "125             spaciou room  0.258202\n",
       "348             wonder hotel  0.253383\n",
       "347                  ad bonu  0.253383\n",
       "367              warm welcom  0.251491\n",
       "158                help room  0.250480\n",
       "97                 top notch  0.247152\n",
       "379                 hot item  0.247152\n",
       "61              mani restaur  0.245932\n",
       "82                club loung  0.245511\n",
       "396                lobbi bar  0.244649\n",
       "188              excel hotel  0.244649\n",
       "44                great view  0.244117\n",
       "63                great valu  0.243983\n",
       "52                 bow river  0.241875\n",
       "200                nice stay  0.240372\n",
       "489                fit facil  0.238784\n",
       "478              restaur bar  0.238784\n",
       "470              first class  0.238784\n",
       "483              free upgrad  0.238784\n",
       "211               great food  0.237102\n",
       "220            perfect locat  0.235318\n",
       "215  kensington riversid inn  0.235318\n",
       "213         carriag hous inn  0.235318\n",
       "497                larg walk  0.235318\n",
       "435                chef tabl  0.235318\n",
       "473               club level  0.231404"
      ]
     },
     "execution_count": 151,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Sorted top pmi words for positive reviews\n",
    "pmiposlist_phrase.sort_values('pmi',ascending=0).head(50)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "For top positive phrases in PMI, we can find many phrases which are not previously observed in topKphrases with ground truth case. This might be because as the PMI is not sensitive to frequency if they occur in all of the positive cases (They get same PMI which can seen in above table). So, as these phrases are only occuring in the Positive cases they getting the same and highest PMI values. \n",
    "\n",
    "Interestingly, we can see that this type of similar values in PMI is not observed in neagtive cases which means that these words might also have appeared in few positive cases, which decreases the negative PMI values."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 152,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>word</th>\n",
       "      <th>pmi</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>485</th>\n",
       "      <td>bed bug</td>\n",
       "      <td>1.281873</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>456</th>\n",
       "      <td>bed sheet</td>\n",
       "      <td>1.244132</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>454</th>\n",
       "      <td>secur deposit</td>\n",
       "      <td>1.236752</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>198</th>\n",
       "      <td>key card</td>\n",
       "      <td>1.228384</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>161</th>\n",
       "      <td>non smoke room</td>\n",
       "      <td>1.199396</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>credit card</td>\n",
       "      <td>1.193451</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>121</th>\n",
       "      <td>smoke room</td>\n",
       "      <td>1.127076</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>242</th>\n",
       "      <td>plu side</td>\n",
       "      <td>1.115055</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>92</th>\n",
       "      <td>air condition</td>\n",
       "      <td>1.068928</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>268</th>\n",
       "      <td>bathroom floor</td>\n",
       "      <td>1.061811</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>464</th>\n",
       "      <td>cigarett smoke</td>\n",
       "      <td>1.056534</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>152</th>\n",
       "      <td>second night</td>\n",
       "      <td>1.043462</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>462</th>\n",
       "      <td>toilet seat</td>\n",
       "      <td>1.036493</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>345</th>\n",
       "      <td>front desk person</td>\n",
       "      <td>1.017075</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>451</th>\n",
       "      <td>first thing</td>\n",
       "      <td>1.013609</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>325</th>\n",
       "      <td>cell phone</td>\n",
       "      <td>0.999010</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>356</th>\n",
       "      <td>second room</td>\n",
       "      <td>0.999010</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>383</th>\n",
       "      <td>next room</td>\n",
       "      <td>0.978923</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>155</th>\n",
       "      <td>differ room</td>\n",
       "      <td>0.968011</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>328</th>\n",
       "      <td>small room</td>\n",
       "      <td>0.956450</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>156</th>\n",
       "      <td>ice machin</td>\n",
       "      <td>0.945021</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>406</th>\n",
       "      <td>room door</td>\n",
       "      <td>0.931132</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>183</th>\n",
       "      <td>phone call</td>\n",
       "      <td>0.917229</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>139</th>\n",
       "      <td>desk clerk</td>\n",
       "      <td>0.909930</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>355</th>\n",
       "      <td>room number</td>\n",
       "      <td>0.882342</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>210</th>\n",
       "      <td>first room</td>\n",
       "      <td>0.880077</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>349</th>\n",
       "      <td>hotel manag</td>\n",
       "      <td>0.835822</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>352</th>\n",
       "      <td>car park</td>\n",
       "      <td>0.835822</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>49</th>\n",
       "      <td>hot water</td>\n",
       "      <td>0.831287</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>492</th>\n",
       "      <td>hair dryer</td>\n",
       "      <td>0.831287</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>95</th>\n",
       "      <td>good thing</td>\n",
       "      <td>0.825114</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>453</th>\n",
       "      <td>decent hotel</td>\n",
       "      <td>0.815372</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>119</th>\n",
       "      <td>long time</td>\n",
       "      <td>0.784600</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>181</th>\n",
       "      <td>first day</td>\n",
       "      <td>0.783837</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>281</th>\n",
       "      <td>motel villag</td>\n",
       "      <td>0.782097</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>192</th>\n",
       "      <td>new room</td>\n",
       "      <td>0.774129</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>302</th>\n",
       "      <td>differ hotel</td>\n",
       "      <td>0.774129</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>171</th>\n",
       "      <td>comfort inn</td>\n",
       "      <td>0.768852</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>330</th>\n",
       "      <td>front desk agent</td>\n",
       "      <td>0.755779</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>179</th>\n",
       "      <td>nthe bathroom</td>\n",
       "      <td>0.739386</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>486</th>\n",
       "      <td>nuvo hotel</td>\n",
       "      <td>0.733307</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>222</th>\n",
       "      <td>last day</td>\n",
       "      <td>0.725926</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>398</th>\n",
       "      <td>flat screen tv</td>\n",
       "      <td>0.715288</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>366</th>\n",
       "      <td>hotel stay</td>\n",
       "      <td>0.711328</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>357</th>\n",
       "      <td>2nd floor</td>\n",
       "      <td>0.711328</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>372</th>\n",
       "      <td>bath tub</td>\n",
       "      <td>0.711328</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>279</th>\n",
       "      <td>second day</td>\n",
       "      <td>0.705136</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>park lot</td>\n",
       "      <td>0.696939</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>174</th>\n",
       "      <td>front desk clerk</td>\n",
       "      <td>0.695436</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>458</th>\n",
       "      <td>travel agent</td>\n",
       "      <td>0.668768</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                  word       pmi\n",
       "485            bed bug  1.281873\n",
       "456          bed sheet  1.244132\n",
       "454      secur deposit  1.236752\n",
       "198           key card  1.228384\n",
       "161     non smoke room  1.199396\n",
       "17         credit card  1.193451\n",
       "121         smoke room  1.127076\n",
       "242           plu side  1.115055\n",
       "92       air condition  1.068928\n",
       "268     bathroom floor  1.061811\n",
       "464     cigarett smoke  1.056534\n",
       "152       second night  1.043462\n",
       "462        toilet seat  1.036493\n",
       "345  front desk person  1.017075\n",
       "451        first thing  1.013609\n",
       "325         cell phone  0.999010\n",
       "356        second room  0.999010\n",
       "383          next room  0.978923\n",
       "155        differ room  0.968011\n",
       "328         small room  0.956450\n",
       "156         ice machin  0.945021\n",
       "406          room door  0.931132\n",
       "183         phone call  0.917229\n",
       "139         desk clerk  0.909930\n",
       "355        room number  0.882342\n",
       "210         first room  0.880077\n",
       "349        hotel manag  0.835822\n",
       "352           car park  0.835822\n",
       "49           hot water  0.831287\n",
       "492         hair dryer  0.831287\n",
       "95          good thing  0.825114\n",
       "453       decent hotel  0.815372\n",
       "119          long time  0.784600\n",
       "181          first day  0.783837\n",
       "281       motel villag  0.782097\n",
       "192           new room  0.774129\n",
       "302       differ hotel  0.774129\n",
       "171        comfort inn  0.768852\n",
       "330   front desk agent  0.755779\n",
       "179      nthe bathroom  0.739386\n",
       "486         nuvo hotel  0.733307\n",
       "222           last day  0.725926\n",
       "398     flat screen tv  0.715288\n",
       "366         hotel stay  0.711328\n",
       "357          2nd floor  0.711328\n",
       "372           bath tub  0.711328\n",
       "279         second day  0.705136\n",
       "29            park lot  0.696939\n",
       "174   front desk clerk  0.695436\n",
       "458       travel agent  0.668768"
      ]
     },
     "execution_count": 152,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Sorted top pmi words for negative reviews\n",
    "pmineglist_phrase.sort_values('pmi',ascending=0).head(50)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>ratingScore</th>\n",
       "      <th>review_counts</th>\n",
       "      <th>Vader_score</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>hotelName</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Brother Li Homestay</th>\n",
       "      <td>5.000000</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.957100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Bonaventure Bed and Breakfast</th>\n",
       "      <td>4.937500</td>\n",
       "      <td>16.0</td>\n",
       "      <td>0.952269</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Novel Bed and Breakfast</th>\n",
       "      <td>4.923077</td>\n",
       "      <td>13.0</td>\n",
       "      <td>0.947500</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                               ratingScore  review_counts  Vader_score\n",
       "hotelName                                                             \n",
       "Brother Li Homestay               5.000000            2.0     0.957100\n",
       "Bonaventure Bed and Breakfast     4.937500           16.0     0.952269\n",
       "Novel Bed and Breakfast           4.923077           13.0     0.947500"
      ]
     },
     "execution_count": 56,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "hotels_avg_data.sort_values(by=['ratingScore'], ascending=False).head(3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 187,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_topRatedHotel = finaldf[finaldf['hotelName']== 'Bonaventure Bed and Breakfast']\n",
    "df_topRatedHotel_phrases = finaldf_phrase[finaldf_phrase['hotelName']== 'Bonaventure Bed and Breakfast']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 188,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>hotelName</th>\n",
       "      <th>reviewColumn</th>\n",
       "      <th>ratingScore</th>\n",
       "      <th>groundTruth</th>\n",
       "      <th>Vader_score</th>\n",
       "      <th>hotel</th>\n",
       "      <th>room</th>\n",
       "      <th>staff</th>\n",
       "      <th>stay</th>\n",
       "      <th>great</th>\n",
       "      <th>...</th>\n",
       "      <th>variety</th>\n",
       "      <th>using</th>\n",
       "      <th>real</th>\n",
       "      <th>elevators</th>\n",
       "      <th>maintained</th>\n",
       "      <th>adequate</th>\n",
       "      <th>beyond</th>\n",
       "      <th>choices</th>\n",
       "      <th>nthere</th>\n",
       "      <th>blocks</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>11590</th>\n",
       "      <td>Bonaventure Bed and Breakfast</td>\n",
       "      <td>\"We stayed in this first rate B &amp;amp; B in the...</td>\n",
       "      <td>5</td>\n",
       "      <td>positive</td>\n",
       "      <td>0.9891</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11591</th>\n",
       "      <td>Bonaventure Bed and Breakfast</td>\n",
       "      <td>\"This is the best B&amp;amp;B we have ever stayed ...</td>\n",
       "      <td>5</td>\n",
       "      <td>positive</td>\n",
       "      <td>0.9931</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11592</th>\n",
       "      <td>Bonaventure Bed and Breakfast</td>\n",
       "      <td>\"Bonaventure is a lovely B&amp;amp;B made perfect ...</td>\n",
       "      <td>5</td>\n",
       "      <td>positive</td>\n",
       "      <td>0.9912</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>3 rows × 505 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                           hotelName  \\\n",
       "11590  Bonaventure Bed and Breakfast   \n",
       "11591  Bonaventure Bed and Breakfast   \n",
       "11592  Bonaventure Bed and Breakfast   \n",
       "\n",
       "                                            reviewColumn  ratingScore  \\\n",
       "11590  \"We stayed in this first rate B &amp; B in the...            5   \n",
       "11591  \"This is the best B&amp;B we have ever stayed ...            5   \n",
       "11592  \"Bonaventure is a lovely B&amp;B made perfect ...            5   \n",
       "\n",
       "      groundTruth  Vader_score  hotel  room  staff  stay  great   ...    \\\n",
       "11590    positive       0.9891      0     1      0     0      1   ...     \n",
       "11591    positive       0.9931      0     0      0     1      1   ...     \n",
       "11592    positive       0.9912      0     1      0     0      1   ...     \n",
       "\n",
       "       variety  using  real  elevators  maintained  adequate  beyond  choices  \\\n",
       "11590        0      0     0          0           0         0       0        0   \n",
       "11591        0      0     0          0           0         0       0        0   \n",
       "11592        0      0     0          0           0         0       0        0   \n",
       "\n",
       "       nthere  blocks  \n",
       "11590       0       0  \n",
       "11591       0       0  \n",
       "11592       0       0  \n",
       "\n",
       "[3 rows x 505 columns]"
      ]
     },
     "execution_count": 188,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_topRatedHotel.head(3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 189,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>ratingScore</th>\n",
       "      <th>review_counts</th>\n",
       "      <th>Vader_score</th>\n",
       "      <th>Variance_ratingscore</th>\n",
       "      <th>Variance_Vaderscore</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>hotelName</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Town &amp;amp; Country Motor Hotel</th>\n",
       "      <td>1.300000</td>\n",
       "      <td>10.0</td>\n",
       "      <td>-0.112840</td>\n",
       "      <td>0.455556</td>\n",
       "      <td>0.676837</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Olympia Motel</th>\n",
       "      <td>1.090909</td>\n",
       "      <td>11.0</td>\n",
       "      <td>-0.025009</td>\n",
       "      <td>0.090909</td>\n",
       "      <td>0.513756</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>O Stays Waterfront</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>6.0</td>\n",
       "      <td>0.173583</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.585528</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                ratingScore  review_counts  Vader_score  \\\n",
       "hotelName                                                                 \n",
       "Town &amp; Country Motor Hotel     1.300000           10.0    -0.112840   \n",
       "Olympia Motel                      1.090909           11.0    -0.025009   \n",
       "O Stays Waterfront                 1.000000            6.0     0.173583   \n",
       "\n",
       "                                Variance_ratingscore  Variance_Vaderscore  \n",
       "hotelName                                                                  \n",
       "Town &amp; Country Motor Hotel              0.455556             0.676837  \n",
       "Olympia Motel                               0.090909             0.513756  \n",
       "O Stays Waterfront                          0.000000             0.585528  "
      ]
     },
     "execution_count": 189,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "hotels_avg_data.sort_values(by=['ratingScore'], ascending=False).tail(3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 190,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_leastRatedHotel = finaldf[finaldf['hotelName']== 'Olympia Motel']\n",
    "df_leastRatedHotel_phrases = finaldf_phrase[finaldf_phrase['hotelName']== 'Olympia Motel']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 191,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>hotelName</th>\n",
       "      <th>reviewColumn</th>\n",
       "      <th>ratingScore</th>\n",
       "      <th>groundTruth</th>\n",
       "      <th>Vader_score</th>\n",
       "      <th>hotel</th>\n",
       "      <th>room</th>\n",
       "      <th>staff</th>\n",
       "      <th>stay</th>\n",
       "      <th>great</th>\n",
       "      <th>...</th>\n",
       "      <th>variety</th>\n",
       "      <th>using</th>\n",
       "      <th>real</th>\n",
       "      <th>elevators</th>\n",
       "      <th>maintained</th>\n",
       "      <th>adequate</th>\n",
       "      <th>beyond</th>\n",
       "      <th>choices</th>\n",
       "      <th>nthere</th>\n",
       "      <th>blocks</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>6866</th>\n",
       "      <td>Olympia Motel</td>\n",
       "      <td>\"Was looking online for an overnight stay in C...</td>\n",
       "      <td>1</td>\n",
       "      <td>negative</td>\n",
       "      <td>0.3311</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6867</th>\n",
       "      <td>Olympia Motel</td>\n",
       "      <td>\"The most disgusting hotel I have ever seen, a...</td>\n",
       "      <td>1</td>\n",
       "      <td>negative</td>\n",
       "      <td>-0.5697</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6868</th>\n",
       "      <td>Olympia Motel</td>\n",
       "      <td>\"So I\\\\'m not sure how much the \\\\'revamped an...</td>\n",
       "      <td>1</td>\n",
       "      <td>negative</td>\n",
       "      <td>-0.8739</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>3 rows × 505 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "          hotelName                                       reviewColumn  \\\n",
       "6866  Olympia Motel  \"Was looking online for an overnight stay in C...   \n",
       "6867  Olympia Motel  \"The most disgusting hotel I have ever seen, a...   \n",
       "6868  Olympia Motel  \"So I\\\\'m not sure how much the \\\\'revamped an...   \n",
       "\n",
       "      ratingScore groundTruth  Vader_score  hotel  room  staff  stay  great  \\\n",
       "6866            1    negative       0.3311      0     1      0     1      0   \n",
       "6867            1    negative      -0.5697      1     1      0     1      0   \n",
       "6868            1    negative      -0.8739      0     0      0     0      0   \n",
       "\n",
       "       ...    variety  using  real  elevators  maintained  adequate  beyond  \\\n",
       "6866   ...          0      0     0          0           0         0       0   \n",
       "6867   ...          0      0     0          0           0         0       0   \n",
       "6868   ...          0      0     0          0           0         0       0   \n",
       "\n",
       "      choices  nthere  blocks  \n",
       "6866        0       0       0  \n",
       "6867        0       0       0  \n",
       "6868        0       0       0  \n",
       "\n",
       "[3 rows x 505 columns]"
      ]
     },
     "execution_count": 191,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_leastRatedHotel.head(3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 192,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|████████████████████████████████████████████████████████████████████████████████| 500/500 [00:09<00:00, 52.81it/s]\n"
     ]
    }
   ],
   "source": [
    "pmiposlist_topHotel, pmineglist_topHotel, pmidf = pmiForAllCal(df_topRatedHotel)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 193,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>word</th>\n",
       "      <th>pmi</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>hotel</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>room</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>staff</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>stay</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>great</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    word  pmi\n",
       "0  hotel  0.0\n",
       "1   room  0.0\n",
       "2  staff  0.0\n",
       "3   stay  0.0\n",
       "4  great  0.0"
      ]
     },
     "execution_count": 193,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pmiposlist_topHotel[:5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 194,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>word</th>\n",
       "      <th>pmi</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>hotel</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>room</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>staff</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>stay</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>great</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    word  pmi\n",
       "0  hotel  0.0\n",
       "1   room  0.0\n",
       "2  staff  0.0\n",
       "3   stay  0.0\n",
       "4  great  0.0"
      ]
     },
     "execution_count": 194,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pmineglist_topHotel[:5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 195,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|████████████████████████████████████████████████████████████████████████████████| 500/500 [00:09<00:00, 52.94it/s]\n"
     ]
    }
   ],
   "source": [
    "# pmiposlist_phrase, pmineglist_phrase, pmidf_phrase = pmiForAllCal(finaldf_phrase, topk=topk_phrase)\n",
    "pmiposlist_topHotel, pmineglist_topHotel, pmidf = pmiForAllCal(df_topRatedHotel_phrases, topk=topk_phrase)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 196,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>word</th>\n",
       "      <th>pmi</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>front desk</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>hot tub</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>downtown calgari</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>great locat</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>front desk staff</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "               word  pmi\n",
       "0        front desk  0.0\n",
       "1           hot tub  0.0\n",
       "2  downtown calgari  0.0\n",
       "3       great locat  0.0\n",
       "4  front desk staff  0.0"
      ]
     },
     "execution_count": 196,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pmiposlist_topHotel[:5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 197,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>word</th>\n",
       "      <th>pmi</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>front desk</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>hot tub</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>downtown calgari</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>great locat</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>front desk staff</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "               word  pmi\n",
       "0        front desk  0.0\n",
       "1           hot tub  0.0\n",
       "2  downtown calgari  0.0\n",
       "3       great locat  0.0\n",
       "4  front desk staff  0.0"
      ]
     },
     "execution_count": 197,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pmineglist_topHotel[:5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 198,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|████████████████████████████████████████████████████████████████████████████████| 500/500 [00:09<00:00, 52.17it/s]\n"
     ]
    }
   ],
   "source": [
    "pmiposlist_topHotel, pmineglist_topHotel, pmidf = pmiForAllCal(df_leastRatedHotel)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 199,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>word</th>\n",
       "      <th>pmi</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>hotel</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>room</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>staff</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>stay</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>great</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    word  pmi\n",
       "0  hotel  0.0\n",
       "1   room  0.0\n",
       "2  staff  0.0\n",
       "3   stay  0.0\n",
       "4  great  0.0"
      ]
     },
     "execution_count": 199,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pmiposlist_topHotel[:5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 201,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>word</th>\n",
       "      <th>pmi</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>hotel</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>room</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>staff</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>stay</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>great</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    word  pmi\n",
       "0  hotel  0.0\n",
       "1   room  0.0\n",
       "2  staff  0.0\n",
       "3   stay  0.0\n",
       "4  great  0.0"
      ]
     },
     "execution_count": 201,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pmineglist_topHotel[:5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 202,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|████████████████████████████████████████████████████████████████████████████████| 500/500 [00:09<00:00, 54.54it/s]\n"
     ]
    }
   ],
   "source": [
    "pmiposlist_leastHotel, pmineglist_leastHotel, pmidf = pmiForAllCal(df_leastRatedHotel_phrases, topk=topk_phrase)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 203,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>word</th>\n",
       "      <th>pmi</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>front desk</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>hot tub</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>downtown calgari</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>great locat</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>front desk staff</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "               word  pmi\n",
       "0        front desk  0.0\n",
       "1           hot tub  0.0\n",
       "2  downtown calgari  0.0\n",
       "3       great locat  0.0\n",
       "4  front desk staff  0.0"
      ]
     },
     "execution_count": 203,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pmiposlist_leastHotel[:5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 204,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>word</th>\n",
       "      <th>pmi</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>front desk</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>hot tub</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>downtown calgari</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>great locat</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>front desk staff</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "               word  pmi\n",
       "0        front desk  0.0\n",
       "1           hot tub  0.0\n",
       "2  downtown calgari  0.0\n",
       "3       great locat  0.0\n",
       "4  front desk staff  0.0"
      ]
     },
     "execution_count": 204,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pmineglist_leastHotel[:5]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "we are unable to get any hotel-specific insights by considering top and least rated hotels because, In both the cases we have only either positive or negative cases only. Hence,\n",
    "\n",
    "**Top Hotel - All Positive Reviews only:**\n",
    "\n",
    "Case 1 (PMI_positive_list): In this case, when we are calculating PMI_positive_list then as P(Y) = 1 [ probability of class \"Positive\" =1 ] and P(X,Y) & P(X) are same (because probability of occurance of word is same as probability of occurance of word in class positive) and hence PMI will be '0' for each word in this case.\n",
    "\n",
    "Case 2: (PMI_negative_list): In this case, since P(Y) = 0 ([ probability of class \"negative\" = 0 ]) and P(X,Y) will be zero and hence in the calculation a small correction of 0.0001 is added (To avoid floating division error) and we get PMIs as zero since all the terms are near zeros.\n",
    "\n",
    "Similarly when we calculate for the least rated hotel, as all the reviews are zeros we are getting zero PMI values in all the cases."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Q5. General Plots"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [],
   "source": [
    "def getHistogram(measure, title):\n",
    "    if measure=='both':\n",
    "        x = [finaldf['ratingScore'].values/5]\n",
    "        y = [finaldf['Vader_score'].values]\n",
    "        bins = np.linspace(-1, 1, 100)\n",
    "        plt.title(title)\n",
    "        plt.hist(x, bins, label='x')\n",
    "        plt.hist(y, bins, label='y')\n",
    "        plt.legend(loc='upper right')\n",
    "        plt.show()\n",
    "    \n",
    "    else:\n",
    "        plt.hist(finaldf[measure].values)\n",
    "        plt.title(title)\n",
    "        plt.xlabel(\"Value\")\n",
    "        plt.ylabel(\"Frequency\")\n",
    "        fig = plt.gcf()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "getHistogram('ratingScore', 'Ground Truth')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "getHistogram('Vader_score', 'Ground Truth')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "getHistogram('both', 'Ground Truth v.s Vader')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 207,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1280"
      ]
     },
     "execution_count": 207,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(hotelDf[(hotelDf['Vader_score'] < 0)])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 212,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "3249"
      ]
     },
     "execution_count": 212,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(hotelDf[(hotelDf['groundTruth'] == 'negative')])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 214,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "3249"
      ]
     },
     "execution_count": 214,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(hotelDf[(hotelDf['ratingScore'] <= 3)])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Above plots show that distribution for vader and rating scores is almost similar. One Interesting observation from the plot is that, we have very few ratings with vader_scores less than zero although there are many negative reviews (Ground truth)\n",
    "\n",
    "\n",
    "Since ratingScore <= 3 (corresponding to negative reviews), we have very few reviews where the vader predicted compound scores is less zero. This shows that vader is unable to predict these negative reviews. This might be because of the length the of reviews or the complexity of words in the reviews which makes vader to predict them as slightly positive reviews.\n",
    "\n",
    "Also, another possibilty is that as we are considering ratingScore <= 3 as negatives in ground truth, In these cases customer might be acutaully neutral or slightly positive (as vader is predicting)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 215,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYIAAAEWCAYAAABrDZDcAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMi4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvhp/UCwAAGQhJREFUeJzt3XmYZGWZpvH7YVFkcUEKRQRKkXYdKbBckFYRNwRcLzfaBaZRdAa7dXRG0OlBdNwbQZ0eURwUtV0Q3GixVUAFaR2wwBJZRFBLQRAKRUGwpcG3/zhfYnSamRVZVZFRkef+XVdcefbznqioeOJ8J+I7qSokSf210bgLkCSNl0EgST1nEEhSzxkEktRzBoEk9ZxBIEk9ZxBoZJKsSvLEMe37HknOSnJjknePeF87Jvldko1HuZ9xSHJQkrMX2770H20y7gKkETkEuA64c434xzJV9XNgy1HuYxIlORK4X1W9aNy1aG6eEWiDl2RtPrDsBFw8bAgsxk/z87WWz7MWAYOgZ1pzzX9PckGS3yY5Mclmbd6fnZonqST3a8MnJHl/kn9uTSH/kuSeSd6T5PokP0yy27RdPjzJxW3+R6b21ba3f5KVSX6T5NtJHjqtzsOSXADcNNObVJJHJ/luO47vJnn0VJ3AgcDrWp1/1jzVjuXYJF9OchPw+CR3THJUkp8nuSbJB5LcqS1/SZL9B9bfJMl1SXZPsrQ9T5u0eXdJcnySq5P8IslbpoImyc+SPKwNv6it96A2/tIkX2jDj0iyIskNrZajZ/n33CvJlUne0OpZleSFA/PnOqapdQ9L8kvgIzPtoy17VPs3/GmSpw5Mv1eSU5L8OsnlSV7Wpu8DvAF4fvs3+P6anhuNj0HQT88D9gHuAzwUOGie6/4dsA3wB+A7wPlt/GRg+hvWC4GnADsDf9HWJcnuwIeBlwN3Bz4InJLkjgPrHgDsB9y1qm4d3GiSrYFTgfe19Y8GTk1y96o6CPgE8K6q2rKqTp/lWP4KeCuwFXA28M5W4zLgfsD2wBFt2U+1eqY8Bbiuqs6fYbsfBW5t29gNeDLw0jbvTGCvNvxY4CfA4wbGz2zD7wXeW1V3pnvuPjPLMQDck+75354uAI9Lcv82b65jmlp3a7ozqENm2f4jgUvbPt4FHJ8kbd6ngCuBewHPAd6W5AlV9RXgbcCJ7d9g1yGeG41LVfno0QNYBbxoYPxdwAfa8EHA2dOWL7p2XoATgA8NzPsb4JKB8f8E/Gbavl4xML4v8OM2fCzwv6ft61LgcQPr/vUcx/Fi4Nxp074DHDRQ61vmWP8E4GMD4wFuAnYemLYH8NM2fD/gRmDzNv4J4Ig2vLQ9T5sA96ALyDsNbOcA4Btt+GDglDZ8Cd2b4Kfb+M+A3dvwWcCbgG3W8O+5F90b6xYD0z4D/K8hjmkv4BZgszm2fxBw+cD45u1Y7wnsANwGbDUw/+3ACW34SOAfB+at6bk5iGmvPx8L87BNsJ9+OTB8M92nuWFdMzD8+xnGp180vWJg+GcD+9oJODDJ3wzMv8O0WgbXne5ebXuDfkb3iXdYg9tfQvcmd96fPuwSYGOAqro8ySXA05L8E/B0uk+00+0EbApcPbCdjQb2dSZwVJJ7tm2fCLwxyVLgLsDKttzBwJuBHyb5KfCmqvrSLMdxfVXdNDA+9TzPeUzN6qr611m2O+X210tV3dy2tSXdmdivq+rGaftePst21vTcaEwMAg26ie6NA4D2ZrWudhgY3hG4qg1fAby1qt46x7pzXei9iu6NZdCOwFfmUdvg9q+jC7IHV9UvZll+qnloI7oL0ZfPsMwVdJ96t6lpzVlwe6DcDPwtcFZV3dja5w+h+zT8x7bcZcABSTYCng2c3Jq9bpq+TeBuSbYYmLcjcOGQx7Qu36i6Ctg6yVYDYbAjMLWv6due87nR+HiNQIO+Dzw4ybJ2UffI9bDNQ5Pcu7Xpv4HuEzDAh4BXJHlkOlsk2S/JVkNu98vAXyT5q3bh9vnAg4DZPjXPqb0Bfwg4Jsm2AEm2T/KUgcU+Tdem/V+AT86ynauBrwHvTnLnJBsl2TnJ4wYWOxN4JX+6HvDNaeNTF5KXtLp+0ybfNschvCnJHZI8BtgfOGnIY1prVXUF8G3g7Uk2S3ex/2C6ZjPozhaXtjAb9rnRGBgEul1V/YiuOeJ04DK6C6jr6pN0//l/0h5vaftaAbwM+AfgeuBy5nHRuqp+RfeG91rgV8DrgP2r6rp1qPWwVsf/T3ID3fMwddF16o3sO8Cj+VOgzeQldM1cF9Md28nAdgPzz6S7QH3WLOPQXcy/KMnv6C4cv2COJpxftv1cRfcm/Iqq+uEwx7QeHEB3jeQq4PPAG6vqtDbvpPb3V0mmLqqv6bnRGKRdpJE0gZLsRXdB9t7jrkWTyzMCSeo5g0CSes6mIUnqOc8IJKnnJuJ3BNtss00tXbp03GVI0kQ577zzrquqJWtabiKCYOnSpaxYsWLcZUjSREky/df3M7JpSJJ6ziCQpJ4zCCSp5wwCSeo5g0CSes4gkKSeG1kQJNkhyTfS3ev1oiSvatOPbPcqXdke+46qBknSmo3ydwS3Aq+tqvNbH/PnJZnqnvaYqjpqhPuWJA1pZEHQ+m6/ug3f2G7zN5/bCEqSFsCC/LK43Y91N+AcYE/glUleAqygO2u4foZ1DqG7fR877rjjWu976eGnrvW662rVO/Yb274laVgjv1icZEvgs8Crq+oG4FhgZ2AZ3RnDu2dar6qOq6rlVbV8yZI1dpUhSVpLIw2CJJvShcAnqupzAFV1TVXdNnA/1UeMsgZJ0txG+a2hAMcDl1TV0QPTB+9P+izgwlHVIElas1FeI9gTeDHwgyQr27Q3AAckWQYUsAp4+QhrkCStwSi/NXQ2kBlmfXlU+5QkzZ+/LJaknjMIJKnnDAJJ6jmDQJJ6ziCQpJ4zCCSp5wwCSeo5g0CSes4gkKSeMwgkqecMAknqOYNAknrOIJCknjMIJKnnDAJJ6jmDQJJ6ziCQpJ4zCCSp5wwCSeo5g0CSes4gkKSeMwgkqecMAknqOYNAknrOIJCknjMIJKnnDAJJ6jmDQJJ6ziCQpJ4zCCSp5wwCSeo5g0CSem5kQZBkhyTfSHJJkouSvKpN3zrJaUkua3/vNqoaJElrNsozgluB11bVA4FHAYcmeRBwOHBGVe0CnNHGJUljMrIgqKqrq+r8NnwjcAmwPfAM4KNtsY8CzxxVDZKkNVuQawRJlgK7AecA96iqq6ELC2DbWdY5JMmKJCtWr169EGVKUi+NPAiSbAl8Fnh1Vd0w7HpVdVxVLa+q5UuWLBldgZLUcyMNgiSb0oXAJ6rqc23yNUm2a/O3A64dZQ2SpLmN8ltDAY4HLqmqowdmnQIc2IYPBL44qhokSWu2yQi3vSfwYuAHSVa2aW8A3gF8JsnBwM+B546wBknSGowsCKrqbCCzzH7CqPYrSZoff1ksST1nEEhSzxkEktRzBoEk9ZxBIEk9ZxBIUs8ZBJLUcwaBJPWcQSBJPWcQSFLPGQSS1HMGgST1nEEgST1nEEhSzxkEktRzBoEk9ZxBIEk9ZxBIUs8ZBJLUcwaBJPWcQSBJPWcQSFLPGQSS1HMGgST1nEEgST1nEEhSzxkEktRzmwyzUJKHVNWFoy5msVl6+Klj2e+qd+w3lv1KmkzDnhF8IMm5Sf5rkruOtCJJ0oIaKgiq6i+BFwI7ACuSfDLJk0ZamSRpQQx9jaCqLgP+DjgMeBzwviQ/TPLsURUnSRq9oYIgyUOTHANcAuwNPK2qHtiGjxlhfZKkERv2jOAfgPOBXavq0Ko6H6CqrqI7S/gzST6c5NokFw5MOzLJL5KsbI991/UAJEnrZqhvDQH7Ar+vqtsAkmwEbFZVN1fVx2dZ5wS6APnYtOnHVNVRa1OsJGn9G/aM4HTgTgPjm7dps6qqs4Bfr2VdkqQFMmwQbFZVv5saacObr+U+X5nkgtZ0dLe13IYkaT0ZNghuSrL71EiShwG/X4v9HQvsDCwDrgbePduCSQ5JsiLJitWrV6/FriRJwxj2GsGrgZOSXNXGtwOeP9+dVdU1U8NJPgR8aY5ljwOOA1i+fHnNd1+SpOEMFQRV9d0kDwDuDwT4YVX923x3lmS7qrq6jT4LsNsKSRqzYc8IAB4OLG3r7JaEqpr+jaDbJfkUsBewTZIrgTcCeyVZBhSwCnj52pUtSVpfhu107uN0bfsrgdva5OLPvxp6u6o6YIbJx8+3QEnSaA17RrAceFBV2VYvSYvMsN8auhC45ygLkSSNx7BnBNsAFyc5F/jD1MSqevpIqpIkLZhhg+DIURYhSRqfYb8+emaSnYBdqur0JJsDG4+2NEnSQhi2G+qXAScDH2yTtge+MKqiJEkLZ9iLxYcCewI3wO03qdl2VEVJkhbOsEHwh6q6ZWokySZ0vyOQJE24YYPgzCRvAO7U7lV8EvBPoytLkrRQhg2Cw4HVwA/ouoX4MrPcmUySNFmG/dbQH4EPtYckaREZtq+hnzLDNYGquu96r0iStKDm09fQlM2A5wJbr/9yJEkLbahrBFX1q4HHL6rqPcDeI65NkrQAhm0a2n1gdCO6M4StRlKRJGlBDds0NHhv4VvpbirzvPVejSRpwQ37raHHj7oQSdJ4DNs09Jq55lfV0eunHEnSQpvPt4YeDpzSxp8GnAVcMYqiJEkLZz43ptm9qm4ESHIkcFJVvXRUhWnyLD381LHte9U79hvbvqVJN2wXEzsCtwyM3wIsXe/VSJIW3LBnBB8Hzk3yebpfGD8L+NjIqpIkLZhhvzX01iT/DDymTfrPVfW90ZUlSVoowzYNAWwO3FBV7wWuTHKfEdUkSVpAw96q8o3AYcDr26RNgX8cVVGSpIUz7BnBs4CnAzcBVNVV2MWEJC0KwwbBLVVVtK6ok2wxupIkSQtp2CD4TJIPAndN8jLgdLxJjSQtCsN+a+iodq/iG4D7A0dU1WkjrUyStCDWGARJNga+WlVPBHzzl6RFZo1NQ1V1G3BzkrssQD2SpAU27C+L/xX4QZLTaN8cAqiqvx1JVZKkBTNsEJzaHpKkRWbOIEiyY1X9vKo+Ot8NJ/kwsD9wbVU9pE3bGjiRrsO6VcDzqur6+W5bkrT+rOkawRemBpJ8dp7bPgHYZ9q0w4EzqmoX4Iw2LkkaozUFQQaG7zufDVfVWcCvp01+BjB1dvFR4Jnz2aYkaf1bUxDULMNr6x5VdTVA+7vtbAsmOSTJiiQrVq9evR52LUmayZqCYNckNyS5EXhoG74hyY1JbhhlYVV1XFUtr6rlS5YsGeWuJKnX5rxYXFUbr+f9XZNku6q6Osl2wLXrefuSpHmaz/0I1odTgAPb8IHAFxd4/5KkaUYWBEk+BXwHuH+SK5McDLwDeFKSy4AntXFJ0hgN+4OyeauqA2aZ9YRR7VOSNH8L3TQkSdrAGASS1HMjaxrS+Cw93G6hJA3PMwJJ6jmDQJJ6ziCQpJ4zCCSp5wwCSeo5g0CSes4gkKSeMwgkqecMAknqOYNAknrOIJCknjMIJKnnDAJJ6jmDQJJ6ziCQpJ4zCCSp5wwCSeo5g0CSes4gkKSeMwgkqecMAknqOYNAknrOIJCknjMIJKnnDAJJ6jmDQJJ6ziCQpJ4zCCSp5wwCSeq5Tcax0ySrgBuB24Bbq2r5OOqQJI0pCJrHV9V1Y9y/JAmbhiSp98YVBAV8Lcl5SQ6ZaYEkhyRZkWTF6tWrF7g8SeqPcQXBnlW1O/BU4NAkj52+QFUdV1XLq2r5kiVLFr5CSeqJsQRBVV3V/l4LfB54xDjqkCSNIQiSbJFkq6lh4MnAhQtdhySpM45vDd0D+HySqf1/sqq+MoY6JEmMIQiq6ifArgu9X0nSzPz6qCT1nEEgST1nEEhSz42ziwlpvVl6+Klj2e+qd+w3lv2O07iea+jn870QPCOQpJ4zCCSp5wwCSeo5g0CSes4gkKSeMwgkqecMAknqOYNAknrOIJCknjMIJKnnDAJJ6jmDQJJ6ziCQpJ4zCCSp5wwCSeo570cgTahx3hdgXPp4zAtxDwbPCCSp5wwCSeo5g0CSes4gkKSeMwgkqecMAknqOYNAknrOIJCknvMHZdI66OMPnLT4eEYgST1nEEhSzxkEktRzBoEk9dxYgiDJPkkuTXJ5ksPHUYMkqbPgQZBkY+D/Ak8FHgQckORBC12HJKkzjjOCRwCXV9VPquoW4NPAM8ZQhySJ8fyOYHvgioHxK4FHTl8oySHAIW30d0kuXcv9bQNct5brjtsk1w6TXb+1j88k17/ea88712n1nYZZaBxBkBmm1Z9NqDoOOG6dd5asqKrl67qdcZjk2mGy67f28Znk+ie19nE0DV0J7DAwfm/gqjHUIUliPEHwXWCXJPdJcgfgBcApY6hDksQYmoaq6tYkrwS+CmwMfLiqLhrhLte5eWmMJrl2mOz6rX18Jrn+iaw9VX/WPC9J6hF/WSxJPWcQSFLPLdogmIRuLJJ8OMm1SS4cmLZ1ktOSXNb+3q1NT5L3teO5IMnu46sckuyQ5BtJLklyUZJXTUr9STZLcm6S77fa39Sm3yfJOa32E9uXGUhyxzZ+eZu/dFy1T0mycZLvJflSG5+k2lcl+UGSlUlWtGkb/Oum1XPXJCcn+WF77e8xKbXPZVEGwQR1Y3ECsM+0aYcDZ1TVLsAZbRy6Y9mlPQ4Bjl2gGmdzK/Daqnog8Cjg0PYcT0L9fwD2rqpdgWXAPkkeBbwTOKbVfj1wcFv+YOD6qrofcExbbtxeBVwyMD5JtQM8vqqWDXznfhJeNwDvBb5SVQ8AdqX7N5iU2mdXVYvuAewBfHVg/PXA68dd1yy1LgUuHBi/FNiuDW8HXNqGPwgcMNNyG8ID+CLwpEmrH9gcOJ/u1+3XAZtMfw3RfcNtjza8SVsuY6z53nRvOHsDX6L7keZE1N7qWAVsM23aBv+6Ae4M/HT68zcJta/psSjPCJi5G4vtx1TLfN2jqq4GaH+3bdM32GNqzQ27AecwIfW3ppWVwLXAacCPgd9U1a0z1Hd77W3+b4G7L2zF/8F7gNcBf2zjd2dyaoeuJ4GvJTmvdSUDk/G6uS+wGvhIa5b7f0m2YDJqn9NiDYKhurGYMBvkMSXZEvgs8OqqumGuRWeYNrb6q+q2qlpG9+n6EcADZ1qs/d1gak+yP3BtVZ03OHmGRTe42gfsWVW70zWdHJrksXMsuyHVvwmwO3BsVe0G3MSfmoFmsiHVPqfFGgST3I3FNUm2A2h/r23TN7hjSrIpXQh8oqo+1yZPTP0AVfUb4Jt01znummTqR5aD9d1ee5t/F+DXC1vp7fYEnp5kFV3PvXvTnSFMQu0AVNVV7e+1wOfpgngSXjdXAldW1Tlt/GS6YJiE2ue0WINgkruxOAU4sA0fSNf2PjX9Je2bCI8Cfjt1OjoOSQIcD1xSVUcPzNrg60+yJMld2/CdgCfSXfT7BvCcttj02qeO6TnA16s1+i60qnp9Vd27qpbSva6/XlUvZAJqB0iyRZKtpoaBJwMXMgGvm6r6JXBFkvu3SU8ALmYCal+jcV+kGNUD2Bf4EV3b7/8cdz2z1Pgp4Grg3+g+PRxM1357BnBZ+7t1WzZ034T6MfADYPmYa/9LutPcC4CV7bHvJNQPPBT4Xqv9QuCINv2+wLnA5cBJwB3b9M3a+OVt/n3H/dppde0FfGmSam91fr89Lpr6vzkJr5tWzzJgRXvtfAG426TUPtfDLiYkqecWa9OQJGlIBoEk9ZxBIEk9ZxBIUs8ZBJLUcwaBFr0kb0+yV5JnZpaeaJMcmeQXrUfMi5McsA77e3OSJ659xdLCMgjUB4+k6wfpccC35ljumOq6nXgG8MH2y+l5q6ojqur0tVlXGgeDQItWkr9PcgHwcOA7wEuBY5McMdd6VXUZcDPdj4VIsnOSr7RO0r6V5AFJ7tL61d+oLbN5kiuSbJrkhCTPadMfluTMtu5Xk2yXZNsk57X5uyapJDu28R+3bT03yYXp7plw1qieIwnGcPN6aaFU1f9IchLwYuA1wDeras81rdduIHJZdX3hQHdD8ldU1WVJHgm8v6r2TvJ9urOMbwBPo+v6+d+63jdu74vp/wDPqKrVSZ4PvLWq/jrdzXHuDDyG7peqj0lyNl2Hcje3sHpKVf1iqjsMaVQMAi12u9F1f/EAun5h5vLfkryMrhuEfeD23lUfDZw09QYP3LH9PRF4Pl0QvAB4/7Tt3R94CHBaW3djui5FAL5N14HcY4G3tf2FPzVd/QtwQpLPAJ9DGiGDQItSkmV0d4C7N93NWDbvJmcl3Y1afj/DasdU1VFJng18LMnOdM2nv2nXDqY7BXh7kq2BhwFfn14GcFFV7THDut+iOxvYia6TssPo+m76EkBVvaKdfewHrEyyrKp+NfwzIA3PawRalKpqZXvz/hHd7Uq/TtfUsmyWEBhc93N0zTUHVnePhZ8meS7cfh/aXdtyv6PryO29dJ2/3TZtU5cCS5Ls0dbdNMmD27yzgBfRNUH9ka5r6H3pzgRIsnNVnVNVR9AF2Q5II2IQaNFKsoTufr1/BB5QVWtqGhr0ZuA17WLwC4GD2zWBi+i+VTTlRLo39BOnb6CqbqHr+vmdbd2VdM1MVNWqttjUheCz6c48rm/jf5/uBu8XtmW+P4/apXmx91FJ6jnPCCSp5wwCSeo5g0CSes4gkKSeMwgkqecMAknqOYNAknru3wEDZ+6dpCeGiQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.hist(finaldf['hotelName'].value_counts())\n",
    "plt.title('number of reviews per hotel')\n",
    "plt.xlabel(\"# Reviews\")\n",
    "plt.ylabel(\"Frequency\")\n",
    "fig = plt.gcf()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can observe that most of the hotels have reviews less than 200. And very few hotels with very high reviews. (These might be the few famous hotels and hence they are getting more customers/reviews)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 216,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>ratingScore</th>\n",
       "      <th>review_counts</th>\n",
       "      <th>Vader_score</th>\n",
       "      <th>Variance_ratingscore</th>\n",
       "      <th>Variance_Vaderscore</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>hotelName</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Brother Li Homestay</th>\n",
       "      <td>5.000000</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.957100</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000063</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Bonaventure Bed and Breakfast</th>\n",
       "      <td>4.937500</td>\n",
       "      <td>16.0</td>\n",
       "      <td>0.952269</td>\n",
       "      <td>0.062500</td>\n",
       "      <td>0.003378</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Novel Bed and Breakfast</th>\n",
       "      <td>4.923077</td>\n",
       "      <td>13.0</td>\n",
       "      <td>0.947500</td>\n",
       "      <td>0.076923</td>\n",
       "      <td>0.002905</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Along River Ridge</th>\n",
       "      <td>4.875000</td>\n",
       "      <td>56.0</td>\n",
       "      <td>0.945861</td>\n",
       "      <td>0.111364</td>\n",
       "      <td>0.010659</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>English Bed, Breakfast and Day Spa</th>\n",
       "      <td>4.848485</td>\n",
       "      <td>33.0</td>\n",
       "      <td>0.886221</td>\n",
       "      <td>0.320076</td>\n",
       "      <td>0.117906</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                    ratingScore  review_counts  Vader_score  \\\n",
       "hotelName                                                                     \n",
       "Brother Li Homestay                    5.000000            2.0     0.957100   \n",
       "Bonaventure Bed and Breakfast          4.937500           16.0     0.952269   \n",
       "Novel Bed and Breakfast                4.923077           13.0     0.947500   \n",
       "Along River Ridge                      4.875000           56.0     0.945861   \n",
       "English Bed, Breakfast and Day Spa     4.848485           33.0     0.886221   \n",
       "\n",
       "                                    Variance_ratingscore  Variance_Vaderscore  \n",
       "hotelName                                                                      \n",
       "Brother Li Homestay                             0.000000             0.000063  \n",
       "Bonaventure Bed and Breakfast                   0.062500             0.003378  \n",
       "Novel Bed and Breakfast                         0.076923             0.002905  \n",
       "Along River Ridge                               0.111364             0.010659  \n",
       "English Bed, Breakfast and Day Spa              0.320076             0.117906  "
      ]
     },
     "execution_count": 216,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tp5gthotel = hotels_avg_data.sort_values(by=['ratingScore'], ascending=False).head()\n",
    "# tp5gthotel['hotelName'].values\n",
    "tp5gthotel"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Box plots"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 217,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#Plot top 5 side-by-side boxplot for top 5 ground truth rated hotel\n",
    "tp5gthotel = hotels_avg_data.sort_values(by=['ratingScore'], ascending=False).head()\n",
    "tp5gthotel = tp5gthotel.reset_index()\n",
    "tp5gthotel['hotelName'].values\n",
    "\n",
    "hampton = finaldf.loc[finaldf['hotelName'] == tp5gthotel['hotelName'].values[0]]['ratingScore']\n",
    "towneplace = finaldf.loc[finaldf['hotelName'] == tp5gthotel['hotelName'].values[1]]['ratingScore']\n",
    "hiltonGarden = finaldf.loc[finaldf['hotelName'] == tp5gthotel['hotelName'].values[2]]['ratingScore']\n",
    "homewood = finaldf.loc[finaldf['hotelName'] == tp5gthotel['hotelName'].values[3]]['ratingScore']\n",
    "courtyard = finaldf.loc[finaldf['hotelName'] == tp5gthotel['hotelName'].values[4]]['ratingScore']\n",
    "\n",
    "data = [hampton, towneplace, hiltonGarden, homewood, courtyard]\n",
    "# multiple box plots on one figure\n",
    "plt.figure()\n",
    "plt.boxplot(data)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In this box plots for top 5 hotels using ground truth star rating we can observe that all these hotels have median ratings as 5. Since in our case all the top 5 hotels have almost rating 5 in all the reviews except 1 or 2 reviews which has rating 4, hence clearly the median is 5 and the ratings 4 are clear outliers in this case with out any other spread in the data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#Plot top 5 side-by-side boxplot for top 5 ground truth rated hotel\n",
    "tp5gthotel = hotels_avg_data.sort_values(by=['ratingScore'], ascending=False).head(5)\n",
    "tp5gthotel = tp5gthotel.reset_index()\n",
    "tp5gthotel['hotelName'].values\n",
    "\n",
    "hampton = finaldf.loc[finaldf['hotelName'] == tp5gthotel['hotelName'].values[0]]['Vader_score']\n",
    "towneplace = finaldf.loc[finaldf['hotelName'] == tp5gthotel['hotelName'].values[1]]['Vader_score']\n",
    "hiltonGarden = finaldf.loc[finaldf['hotelName'] == tp5gthotel['hotelName'].values[2]]['Vader_score']\n",
    "homewood = finaldf.loc[finaldf['hotelName'] == tp5gthotel['hotelName'].values[3]]['Vader_score']\n",
    "courtyard = finaldf.loc[finaldf['hotelName'] == tp5gthotel['hotelName'].values[4]]['Vader_score']\n",
    "\n",
    "data = [hampton, towneplace, hiltonGarden, homewood, courtyard]\n",
    "# multiple box plots on one figure\n",
    "plt.figure()\n",
    "plt.boxplot(data)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "For all the top 5 hotels, we can clearly see that Vader is predicting higher values close to 1. In case of last hotel we can see the spread of scores is very high with many outliers. It is clearly seen that although since there is one review with rating 2 here vader is predicting negative values for them close to -1. Similarly for hotels 3 & 4, since they have more reviews we can see there is a clear spread for vader predicted scores. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#Plot top 5 side-by-side boxplot for top 5 ground truth rated hotel\n",
    "tp5gthotel = hotels_avg_data.sort_values(by=['Vader_score'], ascending=False).head(5)\n",
    "tp5gthotel = tp5gthotel.reset_index()\n",
    "tp5gthotel['hotelName'].values\n",
    "\n",
    "hampton = finaldf.loc[finaldf['hotelName'] == tp5gthotel['hotelName'].values[0]]['Vader_score']\n",
    "towneplace = finaldf.loc[finaldf['hotelName'] == tp5gthotel['hotelName'].values[1]]['Vader_score']\n",
    "hiltonGarden = finaldf.loc[finaldf['hotelName'] == tp5gthotel['hotelName'].values[2]]['Vader_score']\n",
    "homewood = finaldf.loc[finaldf['hotelName'] == tp5gthotel['hotelName'].values[3]]['Vader_score']\n",
    "courtyard = finaldf.loc[finaldf['hotelName'] == tp5gthotel['hotelName'].values[4]]['Vader_score']\n",
    "\n",
    "data = [hampton, towneplace, hiltonGarden, homewood, courtyard]\n",
    "# multiple box plots on one figure\n",
    "plt.figure()\n",
    "plt.boxplot(data)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "For all the top 5 hotels with vaderscores, we can clearly see that Vader is predicting values is close to 1. In case of last hotel we can see the spread of scores is from 0.85 to 1.0 . Comparing only the medians for all the scores we can see that vader is making good median predictions in the range of 0.96-0.97 for these reviews. (Actual avg reviews ratings is close to 5)."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Mean and Variance of the ground truth and Vader sentiment scores for the top-5 ranked hotels according to star rating"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [],
   "source": [
    "hotels_avg_data['Variance_ratingscore'] = hotelDf.groupby('hotelName').var().ratingScore\n",
    "hotels_avg_data['Variance_Vaderscore'] = hotelDf.groupby('hotelName').var().Vader_score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>ratingScore</th>\n",
       "      <th>review_counts</th>\n",
       "      <th>Vader_score</th>\n",
       "      <th>Variance_ratingscore</th>\n",
       "      <th>Variance_Vaderscore</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>hotelName</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Brother Li Homestay</th>\n",
       "      <td>5.000000</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.957100</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000063</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Bonaventure Bed and Breakfast</th>\n",
       "      <td>4.937500</td>\n",
       "      <td>16.0</td>\n",
       "      <td>0.952269</td>\n",
       "      <td>0.062500</td>\n",
       "      <td>0.003378</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Novel Bed and Breakfast</th>\n",
       "      <td>4.923077</td>\n",
       "      <td>13.0</td>\n",
       "      <td>0.947500</td>\n",
       "      <td>0.076923</td>\n",
       "      <td>0.002905</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Along River Ridge</th>\n",
       "      <td>4.875000</td>\n",
       "      <td>56.0</td>\n",
       "      <td>0.945861</td>\n",
       "      <td>0.111364</td>\n",
       "      <td>0.010659</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>English Bed, Breakfast and Day Spa</th>\n",
       "      <td>4.848485</td>\n",
       "      <td>33.0</td>\n",
       "      <td>0.886221</td>\n",
       "      <td>0.320076</td>\n",
       "      <td>0.117906</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                    ratingScore  review_counts  Vader_score  \\\n",
       "hotelName                                                                     \n",
       "Brother Li Homestay                    5.000000            2.0     0.957100   \n",
       "Bonaventure Bed and Breakfast          4.937500           16.0     0.952269   \n",
       "Novel Bed and Breakfast                4.923077           13.0     0.947500   \n",
       "Along River Ridge                      4.875000           56.0     0.945861   \n",
       "English Bed, Breakfast and Day Spa     4.848485           33.0     0.886221   \n",
       "\n",
       "                                    Variance_ratingscore  Variance_Vaderscore  \n",
       "hotelName                                                                      \n",
       "Brother Li Homestay                             0.000000             0.000063  \n",
       "Bonaventure Bed and Breakfast                   0.062500             0.003378  \n",
       "Novel Bed and Breakfast                         0.076923             0.002905  \n",
       "Along River Ridge                               0.111364             0.010659  \n",
       "English Bed, Breakfast and Day Spa              0.320076             0.117906  "
      ]
     },
     "execution_count": 75,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "hotels_avg_data.sort_values(by=['ratingScore'], ascending=False).head(5)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Which do you \f",
    "nd more informative, the boxplots or the mean and variance, or are they equally informative? Why?"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Box plots are more informative because they give information about the spread of the data (1st quartile/3rd quartile/median information). Since in our case we have skewed data (Review ratings/Vader_scores) variance is not a good measure of the spread. Also, it gives information about the outliers. Overall Boxplots helps to compare spread of vader and rating scores for different hotels in a better way."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Scatterplots and heatmaps"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5,0,'Vader Score')"
      ]
     },
     "execution_count": 76,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "y = finaldf['ratingScore'].values\n",
    "x = finaldf['Vader_score'].values\n",
    "plt.plot(x, y,\"o\")\n",
    "plt.ylabel('Rating')\n",
    "plt.xlabel('Vader Score')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {},
   "outputs": [],
   "source": [
    "from scipy.stats.kde import gaussian_kde\n",
    "\n",
    "k = gaussian_kde(np.vstack([x, y]))\n",
    "xi, yi = np.mgrid[x.min():x.max():x.size**0.5*1j,y.min():y.max():y.size**0.5*1j]\n",
    "zi = k(np.vstack([xi.flatten(), yi.flatten()]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0,0.5,'Rating')"
      ]
     },
     "execution_count": 78,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x576 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "cmap = sns.cubehelix_palette(light=1, as_cmap=True)\n",
    "fig = plt.figure(figsize=(6,8))\n",
    "ax1 = fig.add_subplot(211)\n",
    "ax2 = fig.add_subplot(212)\n",
    "\n",
    "ax1.pcolormesh(xi, yi, np.log10(zi.reshape(xi.shape)), cmap=cmap)\n",
    "ax2.contourf(xi, yi, np.log10(zi.reshape(xi.shape)), cmap=cmap)\n",
    "\n",
    "ax1.set_xlim(x.min(), x.max())\n",
    "ax1.set_ylim(y.min(), y.max())\n",
    "ax2.set_xlim(x.min(), x.max())\n",
    "ax2.set_ylim(y.min(), y.max())\n",
    "\n",
    "ax1.set_xlabel('Vader Score')\n",
    "ax1.set_ylabel('Rating')\n",
    "\n",
    "ax2.set_xlabel('Vader Score')\n",
    "ax2.set_ylabel('Rating')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Do you notice anything\n",
    "interesting? What does this tell you about star ratings vs. Vader sentiment scores?\n",
    "What does this tell you about human ratings and/or Vader sentiment analysis?\n",
    "\n",
    "\n",
    "from the above plots we can see that, In some cases even the rating score is 5 vader is predicting it to be negative.  It might happen that user gave aa review 5 but might have given a small complaint or suggestion which is causing vader to give negative predictions. Also, This might be because of the slightly sarcastic reviews, which vader is getting confused. Similarly this might happen when the review rating score is 1 (Sarcastic reviews)\n",
    "\n",
    "Overall we can see that for rating scores of 4/5 vader is predicting many of them to have compound score of 1.0 (clear positive). May be becasue user might have wrirtten a positive review but his rating might be 4 which is possible in most of the cases.\n",
    "\n",
    "For the clear positive 4/5 and clear negative 1 rating scores vader is giving correct predictions."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>hotelName</th>\n",
       "      <th>reviewColumn</th>\n",
       "      <th>ratingScore</th>\n",
       "      <th>groundTruth</th>\n",
       "      <th>Vader_score</th>\n",
       "      <th>hotel</th>\n",
       "      <th>room</th>\n",
       "      <th>staff</th>\n",
       "      <th>stay</th>\n",
       "      <th>great</th>\n",
       "      <th>...</th>\n",
       "      <th>using</th>\n",
       "      <th>real</th>\n",
       "      <th>elevators</th>\n",
       "      <th>maintained</th>\n",
       "      <th>adequate</th>\n",
       "      <th>beyond</th>\n",
       "      <th>choices</th>\n",
       "      <th>nthere</th>\n",
       "      <th>blocks</th>\n",
       "      <th>len_review</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Calgary Airport Marriott In-Terminal Hotel</td>\n",
       "      <td>\"We were excited to stay at the new Marriott i...</td>\n",
       "      <td>5</td>\n",
       "      <td>positive</td>\n",
       "      <td>0.9830</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>967</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Calgary Airport Marriott In-Terminal Hotel</td>\n",
       "      <td>\"We were not able to make our Victoria connect...</td>\n",
       "      <td>5</td>\n",
       "      <td>positive</td>\n",
       "      <td>0.9412</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>367</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Calgary Airport Marriott In-Terminal Hotel</td>\n",
       "      <td>\"Rooms are very modern looking with great king...</td>\n",
       "      <td>5</td>\n",
       "      <td>positive</td>\n",
       "      <td>0.6409</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>550</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Calgary Airport Marriott In-Terminal Hotel</td>\n",
       "      <td>\"We stayed just one night at the Marriott and ...</td>\n",
       "      <td>4</td>\n",
       "      <td>positive</td>\n",
       "      <td>0.4957</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>397</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Calgary Airport Marriott In-Terminal Hotel</td>\n",
       "      <td>\"The Marriot is across from the old Delta. Bot...</td>\n",
       "      <td>5</td>\n",
       "      <td>positive</td>\n",
       "      <td>0.9735</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>425</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 506 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                    hotelName  \\\n",
       "0  Calgary Airport Marriott In-Terminal Hotel   \n",
       "1  Calgary Airport Marriott In-Terminal Hotel   \n",
       "2  Calgary Airport Marriott In-Terminal Hotel   \n",
       "3  Calgary Airport Marriott In-Terminal Hotel   \n",
       "4  Calgary Airport Marriott In-Terminal Hotel   \n",
       "\n",
       "                                        reviewColumn  ratingScore groundTruth  \\\n",
       "0  \"We were excited to stay at the new Marriott i...            5    positive   \n",
       "1  \"We were not able to make our Victoria connect...            5    positive   \n",
       "2  \"Rooms are very modern looking with great king...            5    positive   \n",
       "3  \"We stayed just one night at the Marriott and ...            4    positive   \n",
       "4  \"The Marriot is across from the old Delta. Bot...            5    positive   \n",
       "\n",
       "   Vader_score  hotel  room  staff  stay  great     ...      using  real  \\\n",
       "0       0.9830      1     1      0     1      1     ...          0     0   \n",
       "1       0.9412      1     0      1     0      0     ...          0     0   \n",
       "2       0.6409      0     1      0     1      1     ...          0     0   \n",
       "3       0.4957      1     1      1     0      0     ...          0     0   \n",
       "4       0.9735      0     0      1     0      1     ...          0     0   \n",
       "\n",
       "   elevators  maintained  adequate  beyond  choices  nthere  blocks  \\\n",
       "0          0           0         0       0        0       0       0   \n",
       "1          0           0         0       1        0       0       0   \n",
       "2          1           0         0       0        0       0       0   \n",
       "3          0           0         0       0        0       0       0   \n",
       "4          0           0         0       0        0       0       0   \n",
       "\n",
       "   len_review  \n",
       "0         967  \n",
       "1         367  \n",
       "2         550  \n",
       "3         397  \n",
       "4         425  \n",
       "\n",
       "[5 rows x 506 columns]"
      ]
     },
     "execution_count": 80,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "finaldf['len_review'] = finaldf['reviewColumn'].apply(lambda x: len(x))\n",
    "finaldf.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5,0,'Len of review')"
      ]
     },
     "execution_count": 81,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "y = finaldf['ratingScore'].values\n",
    "x = finaldf['len_review'].values\n",
    "plt.plot(x, y,\"o\")\n",
    "plt.ylabel('Rating')\n",
    "plt.xlabel('Len of review')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "When we consider rating score we cannot ientify any clear trends. The distribution of words vs. ratings is almost similar. For one of the very positive review the length is extremely high which can be observed."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {},
   "outputs": [],
   "source": [
    "from scipy.stats.kde import gaussian_kde\n",
    "\n",
    "k = gaussian_kde(np.vstack([x, y]))\n",
    "xi, yi = np.mgrid[x.min():x.max():x.size**0.5*1j,y.min():y.max():y.size**0.5*1j]\n",
    "zi = k(np.vstack([xi.flatten(), yi.flatten()]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0,0.5,'Rating')"
      ]
     },
     "execution_count": 83,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x576 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "cmap = sns.cubehelix_palette(light=1, as_cmap=True)\n",
    "fig = plt.figure(figsize=(6,8))\n",
    "ax1 = fig.add_subplot(211)\n",
    "ax2 = fig.add_subplot(212)\n",
    "\n",
    "ax1.pcolormesh(xi, yi, np.log10(zi.reshape(xi.shape)), cmap=cmap)\n",
    "ax2.contourf(xi, yi, np.log10(zi.reshape(xi.shape)), cmap=cmap)\n",
    "\n",
    "ax1.set_xlim(x.min(), x.max())\n",
    "ax1.set_ylim(y.min(), y.max())\n",
    "ax2.set_xlim(x.min(), x.max())\n",
    "ax2.set_ylim(y.min(), y.max())\n",
    "\n",
    "ax1.set_xlabel('Vader Score')\n",
    "ax1.set_ylabel('Rating')\n",
    "\n",
    "ax2.set_xlabel('Vader Score')\n",
    "ax2.set_ylabel('Rating')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5,0,'Len of review')"
      ]
     },
     "execution_count": 84,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "y = finaldf['Vader_score'].values\n",
    "x = finaldf['len_review'].values\n",
    "plt.plot(x, y,\"o\")\n",
    "plt.ylabel('Rating')\n",
    "plt.xlabel('Len of review')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Although there is not any clear trends we can say that for more positive or more negative reviews Customers are writing very lengthy reviews which can be seen from above plots for the vader scores. Since vader gives continues scores from -1 to 1 depending on the sentiment this is showing us better trends other than descrete rating scores plots."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "metadata": {},
   "outputs": [],
   "source": [
    "from scipy.stats.kde import gaussian_kde\n",
    "\n",
    "k = gaussian_kde(np.vstack([x, y]))\n",
    "xi, yi = np.mgrid[x.min():x.max():x.size**0.5*1j,y.min():y.max():y.size**0.5*1j]\n",
    "zi = k(np.vstack([xi.flatten(), yi.flatten()]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0,0.5,'Rating')"
      ]
     },
     "execution_count": 86,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x576 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "cmap = sns.cubehelix_palette(light=1, as_cmap=True)\n",
    "fig = plt.figure(figsize=(6,8))\n",
    "ax1 = fig.add_subplot(211)\n",
    "ax2 = fig.add_subplot(212)\n",
    "\n",
    "ax1.pcolormesh(xi, yi, np.log10(zi.reshape(xi.shape)), cmap=cmap)\n",
    "ax2.contourf(xi, yi, np.log10(zi.reshape(xi.shape)), cmap=cmap)\n",
    "\n",
    "ax1.set_xlim(x.min(), x.max())\n",
    "ax1.set_ylim(y.min(), y.max())\n",
    "ax2.set_xlim(x.min(), x.max())\n",
    "ax2.set_ylim(y.min(), y.max())\n",
    "\n",
    "ax1.set_xlabel('Vader Score')\n",
    "ax1.set_ylabel('Rating')\n",
    "\n",
    "ax2.set_xlabel('Vader Score')\n",
    "ax2.set_ylabel('Rating')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 117,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5,0,'Number of review')"
      ]
     },
     "execution_count": 117,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "y = hotels_avg_data['Vader_score'].values\n",
    "x = hotels_avg_data['review_counts'].values\n",
    "plt.plot(x, y,\"o\")\n",
    "plt.ylabel('Rating')\n",
    "plt.xlabel('Number of review')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 118,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5,0,'Number of review')"
      ]
     },
     "execution_count": 118,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "y = hotels_avg_data['ratingScore'].values\n",
    "x = hotels_avg_data['review_counts'].values\n",
    "plt.plot(x, y,\"o\")\n",
    "plt.ylabel('Rating')\n",
    "plt.xlabel('Number of review')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Since in this case, each hotel is a single point on the scatterplot, just by comparing above plot for vader and rating scores we can say that vader is giving good predictions because the disribution of points are almost similar. One interesting observation is that for the hotels with avg.ratings around 1&2 vaders prediction is close to zero but not -1. So, for most of the clear positive reviews vader is also predicting good, but in cases where ratings is 1, the reviews might be sarcastic and hence vader might have predicting them close to zeros."
   ]
  },
  {
   "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
}