1.Webscraping_Early_EDA_and_Cleaning.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "Collapsed": "false"
   },
   "source": [
    "# 1. Webscraping, Early EDA and Cleaning"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "Collapsed": "false"
   },
   "source": [
    "## 1.1 Webscrape Process Explained"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "Collapsed": "false"
   },
   "source": [
    "I set a goal of a minimum of 5000 reviews to scrape, before choosing the specific hotels. I then chose the 5 Hilton hotels with the highest number of reviews, to scrape; London Gatwick Airport, London Metropole, London Euston, London Croydon, and London - West End. \n",
    "Between these 5 hotels there were 17538 reviews, I had plenty room to filter or drop reviews and retain at least my minimum of 5000."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "Collapsed": "false"
   },
   "source": [
    "The structure of each review consisted of a 1-5 scale score rating in bubble form, a review summary, and a detailed review split into p1 and p2 (depending on if there was a read more option). Each page on tripadvisor had 5 reviews per page, so I had to navigate between pages using tripadvisor's next page function. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "Collapsed": "false"
   },
   "source": [
    "The root URL I used was 'www.tripadvisor.co.uk'\n",
    "\n",
    "The 5 starting URL extensions I used were:\n",
    "- '/Hotel_Review-g187051-d239658-Reviews-Hotel_Hilton_London_Gatwick_Airport-Crawley_West_Sussex_England.html/'\n",
    "- '/Hotel_Review-g186338-d193089-Reviews-Hilton_London_Metropole-London_England.html/'\n",
    "- '/Hotel_Review-g186338-d192048-Reviews-Hilton_London_Euston-London_England.html/'\n",
    "- '/Hotel_Review-g186338-d193102-Reviews-DoubleTree_by_Hilton_Hotel_London_West_End-London_England.html/'\n",
    "- '/Hotel_Review-g504167-d192599-Reviews-Hilton_London_Croydon-Croydon_Greater_London_England.html'\n",
    "\n",
    "From these pages I chose to extract 5 different features:\n",
    "- hotel_name\n",
    "- review_summary\n",
    "- review_p1\n",
    "- review_p2\n",
    "- score\n",
    "\n",
    "I used a scrapy spider to crawl the website to scrape the requested data. Scrapy proved the be efficient and fast at extracting the data. I ran the spider script (hotels.py) for around 20 minutes, on the 13th May 2020. I have the scrapy .py files located in Tripadvisor_Webscrape folder."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "Collapsed": "false"
   },
   "source": [
    "## 1.2 Imports"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "Collapsed": "false"
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "Collapsed": "false"
   },
   "outputs": [],
   "source": [
    "sns.set_context('poster')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "Collapsed": "false"
   },
   "outputs": [],
   "source": [
    "df = pd.read_csv('Data/1.tripadvisor_scraped_hotel_reviews.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "Collapsed": "false"
   },
   "source": [
    "## 1.3 Checking for Nulls"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "Collapsed": "false"
   },
   "source": [
    "The initial shape of the dataset was (35078,5). The 5 columns was as expected, but there were double the number of rows as the number of reviews scraped. There were null rows with only hotel_name and no other values, so I removed those rows, bringing us back to the expected 17538."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "Collapsed": "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>hotel_name</th>\n",
       "      <th>review_summary</th>\n",
       "      <th>review_p1</th>\n",
       "      <th>review_p2</th>\n",
       "      <th>score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>Hotel Hilton London Gatwick Airport</td>\n",
       "      <td>Awful...</td>\n",
       "      <td>Its hard to find words that can describe this ...</td>\n",
       "      <td>its just all a bit 'run down'  Poor food, not...</td>\n",
       "      <td>&lt;span class=\"ui_bubble_rating bubble_10\"&gt;&lt;/span&gt;</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>Hotel Hilton London Gatwick Airport</td>\n",
       "      <td>Great start to hokidat</td>\n",
       "      <td>Staying in a hotel that’s part of the airport ...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>&lt;span class=\"ui_bubble_rating bubble_50\"&gt;&lt;/span&gt;</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>Hotel Hilton London Gatwick Airport</td>\n",
       "      <td>A good overnight stay!</td>\n",
       "      <td>My wife and I stayed at this hotel for one nig...</td>\n",
       "      <td>flat-screen TV (with a huge selection of TV c...</td>\n",
       "      <td>&lt;span class=\"ui_bubble_rating bubble_50\"&gt;&lt;/span&gt;</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>Hotel Hilton London Gatwick Airport</td>\n",
       "      <td>Satisfactory one night stay</td>\n",
       "      <td>One night  stay which turned out to be much be...</td>\n",
       "      <td>so higher prices are expected. Overall a very...</td>\n",
       "      <td>&lt;span class=\"ui_bubble_rating bubble_40\"&gt;&lt;/span&gt;</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>Hotel Hilton London Gatwick Airport</td>\n",
       "      <td>Still nice but looking a bit sad, old and dated</td>\n",
       "      <td>We had an executive room and even these are st...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>&lt;span class=\"ui_bubble_rating bubble_40\"&gt;&lt;/span&gt;</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>5</td>\n",
       "      <td>Hotel Hilton London Gatwick Airport</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>6</td>\n",
       "      <td>Hotel Hilton London Gatwick Airport</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>7</td>\n",
       "      <td>Hotel Hilton London Gatwick Airport</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>8</td>\n",
       "      <td>Hotel Hilton London Gatwick Airport</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>9</td>\n",
       "      <td>Hotel Hilton London Gatwick Airport</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                            hotel_name  \\\n",
       "0  Hotel Hilton London Gatwick Airport   \n",
       "1  Hotel Hilton London Gatwick Airport   \n",
       "2  Hotel Hilton London Gatwick Airport   \n",
       "3  Hotel Hilton London Gatwick Airport   \n",
       "4  Hotel Hilton London Gatwick Airport   \n",
       "5  Hotel Hilton London Gatwick Airport   \n",
       "6  Hotel Hilton London Gatwick Airport   \n",
       "7  Hotel Hilton London Gatwick Airport   \n",
       "8  Hotel Hilton London Gatwick Airport   \n",
       "9  Hotel Hilton London Gatwick Airport   \n",
       "\n",
       "                                    review_summary  \\\n",
       "0                                         Awful...   \n",
       "1                           Great start to hokidat   \n",
       "2                           A good overnight stay!   \n",
       "3                      Satisfactory one night stay   \n",
       "4  Still nice but looking a bit sad, old and dated   \n",
       "5                                              NaN   \n",
       "6                                              NaN   \n",
       "7                                              NaN   \n",
       "8                                              NaN   \n",
       "9                                              NaN   \n",
       "\n",
       "                                           review_p1  \\\n",
       "0  Its hard to find words that can describe this ...   \n",
       "1  Staying in a hotel that’s part of the airport ...   \n",
       "2  My wife and I stayed at this hotel for one nig...   \n",
       "3  One night  stay which turned out to be much be...   \n",
       "4  We had an executive room and even these are st...   \n",
       "5                                                NaN   \n",
       "6                                                NaN   \n",
       "7                                                NaN   \n",
       "8                                                NaN   \n",
       "9                                                NaN   \n",
       "\n",
       "                                           review_p2  \\\n",
       "0   its just all a bit 'run down'  Poor food, not...   \n",
       "1                                                NaN   \n",
       "2   flat-screen TV (with a huge selection of TV c...   \n",
       "3   so higher prices are expected. Overall a very...   \n",
       "4                                                NaN   \n",
       "5                                                NaN   \n",
       "6                                                NaN   \n",
       "7                                                NaN   \n",
       "8                                                NaN   \n",
       "9                                                NaN   \n",
       "\n",
       "                                              score  \n",
       "0  <span class=\"ui_bubble_rating bubble_10\"></span>  \n",
       "1  <span class=\"ui_bubble_rating bubble_50\"></span>  \n",
       "2  <span class=\"ui_bubble_rating bubble_50\"></span>  \n",
       "3  <span class=\"ui_bubble_rating bubble_40\"></span>  \n",
       "4  <span class=\"ui_bubble_rating bubble_40\"></span>  \n",
       "5                                               NaN  \n",
       "6                                               NaN  \n",
       "7                                               NaN  \n",
       "8                                               NaN  \n",
       "9                                               NaN  "
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "Collapsed": "false"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(35078, 5)"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "Collapsed": "false"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 35078 entries, 0 to 35077\n",
      "Data columns (total 5 columns):\n",
      "hotel_name        35078 non-null object\n",
      "review_summary    17538 non-null object\n",
      "review_p1         17538 non-null object\n",
      "review_p2         4698 non-null object\n",
      "score             17538 non-null object\n",
      "dtypes: object(5)\n",
      "memory usage: 1.3+ MB\n"
     ]
    }
   ],
   "source": [
    "df.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "Collapsed": "false"
   },
   "outputs": [],
   "source": [
    "df.dropna(subset=['review_p1'],axis=0,inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "Collapsed": "false"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Int64Index: 17538 entries, 0 to 35072\n",
      "Data columns (total 5 columns):\n",
      "hotel_name        17538 non-null object\n",
      "review_summary    17538 non-null object\n",
      "review_p1         17538 non-null object\n",
      "review_p2         4698 non-null object\n",
      "score             17538 non-null object\n",
      "dtypes: object(5)\n",
      "memory usage: 822.1+ KB\n"
     ]
    }
   ],
   "source": [
    "df.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "Collapsed": "false"
   },
   "source": [
    "## 1.4 Converting Score Column"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "Collapsed": "false"
   },
   "outputs": [],
   "source": [
    "df['score'] = df['score'].astype(str).apply(lambda x: x[37]).astype(int)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "Collapsed": "false"
   },
   "outputs": [],
   "source": [
    "df.reset_index(inplace=True,drop=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "Collapsed": "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>hotel_name</th>\n",
       "      <th>review_summary</th>\n",
       "      <th>review_p1</th>\n",
       "      <th>review_p2</th>\n",
       "      <th>score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>Hotel Hilton London Gatwick Airport</td>\n",
       "      <td>Awful...</td>\n",
       "      <td>Its hard to find words that can describe this ...</td>\n",
       "      <td>its just all a bit 'run down'  Poor food, not...</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>Hotel Hilton London Gatwick Airport</td>\n",
       "      <td>Great start to hokidat</td>\n",
       "      <td>Staying in a hotel that’s part of the airport ...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>Hotel Hilton London Gatwick Airport</td>\n",
       "      <td>A good overnight stay!</td>\n",
       "      <td>My wife and I stayed at this hotel for one nig...</td>\n",
       "      <td>flat-screen TV (with a huge selection of TV c...</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>Hotel Hilton London Gatwick Airport</td>\n",
       "      <td>Satisfactory one night stay</td>\n",
       "      <td>One night  stay which turned out to be much be...</td>\n",
       "      <td>so higher prices are expected. Overall a very...</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>Hotel Hilton London Gatwick Airport</td>\n",
       "      <td>Still nice but looking a bit sad, old and dated</td>\n",
       "      <td>We had an executive room and even these are st...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>5</td>\n",
       "      <td>Hilton London Metropole</td>\n",
       "      <td>Great stay</td>\n",
       "      <td>Hilton metropole is centrally located and prov...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>6</td>\n",
       "      <td>Hilton London Metropole</td>\n",
       "      <td>Classic Hilton</td>\n",
       "      <td>1 Thanks for the quick suite upgrade! Love it!...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>7</td>\n",
       "      <td>Hilton London Metropole</td>\n",
       "      <td>Terrible</td>\n",
       "      <td>Absolutely terrible. Cracked ceiling, tiny roo...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>8</td>\n",
       "      <td>Hilton London Metropole</td>\n",
       "      <td>Hilton</td>\n",
       "      <td>Me and my sister stayed at the Hilton metropol...</td>\n",
       "      <td>went through my finger to my elbow. Also when...</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>9</td>\n",
       "      <td>Hilton London Metropole</td>\n",
       "      <td>Clean and tidy. Nice rooms</td>\n",
       "      <td>We stayed here for a weekend away and used it ...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                            hotel_name  \\\n",
       "0  Hotel Hilton London Gatwick Airport   \n",
       "1  Hotel Hilton London Gatwick Airport   \n",
       "2  Hotel Hilton London Gatwick Airport   \n",
       "3  Hotel Hilton London Gatwick Airport   \n",
       "4  Hotel Hilton London Gatwick Airport   \n",
       "5              Hilton London Metropole   \n",
       "6              Hilton London Metropole   \n",
       "7              Hilton London Metropole   \n",
       "8              Hilton London Metropole   \n",
       "9              Hilton London Metropole   \n",
       "\n",
       "                                    review_summary  \\\n",
       "0                                         Awful...   \n",
       "1                           Great start to hokidat   \n",
       "2                           A good overnight stay!   \n",
       "3                      Satisfactory one night stay   \n",
       "4  Still nice but looking a bit sad, old and dated   \n",
       "5                                       Great stay   \n",
       "6                                   Classic Hilton   \n",
       "7                                        Terrible    \n",
       "8                                           Hilton   \n",
       "9                      Clean and tidy. Nice rooms    \n",
       "\n",
       "                                           review_p1  \\\n",
       "0  Its hard to find words that can describe this ...   \n",
       "1  Staying in a hotel that’s part of the airport ...   \n",
       "2  My wife and I stayed at this hotel for one nig...   \n",
       "3  One night  stay which turned out to be much be...   \n",
       "4  We had an executive room and even these are st...   \n",
       "5  Hilton metropole is centrally located and prov...   \n",
       "6  1 Thanks for the quick suite upgrade! Love it!...   \n",
       "7  Absolutely terrible. Cracked ceiling, tiny roo...   \n",
       "8  Me and my sister stayed at the Hilton metropol...   \n",
       "9  We stayed here for a weekend away and used it ...   \n",
       "\n",
       "                                           review_p2  score  \n",
       "0   its just all a bit 'run down'  Poor food, not...      1  \n",
       "1                                                NaN      5  \n",
       "2   flat-screen TV (with a huge selection of TV c...      5  \n",
       "3   so higher prices are expected. Overall a very...      4  \n",
       "4                                                NaN      4  \n",
       "5                                                NaN      4  \n",
       "6                                                NaN      5  \n",
       "7                                                NaN      1  \n",
       "8   went through my finger to my elbow. Also when...      2  \n",
       "9                                                NaN      5  "
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head(10)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "Collapsed": "false"
   },
   "source": [
    "### 1.4.1 Histogram of Scores for Each Hotel"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "Collapsed": "false"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 2592x1440 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(36,20))\n",
    "sns.countplot(x='hotel_name', hue='score',data= df)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "Collapsed": "false"
   },
   "source": [
    "### 1.4.2 Histogram of Scores for All Hotels"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "Collapsed": "false"
   },
   "outputs": [],
   "source": [
    "scores = pd.DataFrame(df.groupby('score').size().sort_values(ascending=False)).reset_index()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "Collapsed": "false"
   },
   "outputs": [],
   "source": [
    "scores.rename(columns={0:'count'},inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "Collapsed": "false"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x864 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(12,12))\n",
    "sns.barplot(x='score',y='count',data=scores)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "Collapsed": "false"
   },
   "source": [
    "## 1.5 Adjusting Class Imbalance for Scores"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "Collapsed": "false"
   },
   "source": [
    "This project entailed the use of classification models, and for reliable results, I had to remove reviews to undo class imbalance. Using this visualisation I saw that were much less reviews with a score of 1 compared to reviews with a score of 3, 4, and 5. To combat this imbalance, I randomly removed reviews with scores of 2, 3, 4, and 5, to match with 1 (1881 reviews). "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "Collapsed": "false"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1881"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "minority_class_len = len(df[df['score'] == 1])\n",
    "minority_class_len"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "Collapsed": "false"
   },
   "outputs": [],
   "source": [
    "score_1_class_indices = df[df['score'] == 1].index\n",
    "score_2_class_indices = df[df['score'] == 2].index\n",
    "score_3_class_indices = df[df['score'] == 3].index\n",
    "score_4_class_indices = df[df['score'] == 4].index\n",
    "score_5_class_indices = df[df['score'] == 5].index"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "Collapsed": "false"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1881\n",
      "1881\n",
      "1881\n",
      "1881\n"
     ]
    }
   ],
   "source": [
    "random_score_2_class_indices = np.random.choice(score_2_class_indices,minority_class_len,replace=False)\n",
    "print(len(random_score_2_class_indices))\n",
    "random_score_3_class_indices = np.random.choice(score_3_class_indices,minority_class_len,replace=False)\n",
    "print(len(random_score_3_class_indices))\n",
    "random_score_4_class_indices = np.random.choice(score_4_class_indices,minority_class_len,replace=False)\n",
    "print(len(random_score_4_class_indices))\n",
    "random_score_5_class_indices = np.random.choice(score_5_class_indices,minority_class_len,replace=False)\n",
    "print(len(random_score_5_class_indices))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "Collapsed": "false"
   },
   "outputs": [],
   "source": [
    "under_sample_indices = np.concatenate([score_1_class_indices,random_score_2_class_indices,random_score_3_class_indices,random_score_4_class_indices,random_score_5_class_indices])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "Collapsed": "false"
   },
   "outputs": [],
   "source": [
    "df = df.loc[under_sample_indices]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "Collapsed": "false"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x864 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(12,12))\n",
    "sns.countplot(x='score',data=df)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "Collapsed": "false"
   },
   "source": [
    "## 1.6 Joining Review Part 1 with Review Part 2 in New Column Review"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "Collapsed": "false"
   },
   "source": [
    "Review part 2 contained the read more part of the review (if there was one), so I decided to join review part 1 with revier part 2 into a single column, review."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "Collapsed": "false"
   },
   "source": [
    "The new data from sources like twitter and reddit will require similar processing (combining parts of reviews), as some will be in one tweet/post, whereas others will be in multiple parts due to the limitations on such platforms."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "Collapsed": "false"
   },
   "outputs": [],
   "source": [
    "df['review_p2'].fillna(value='',inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "Collapsed": "false"
   },
   "outputs": [],
   "source": [
    "df['review'] = df['review_p1'].astype(str) + df['review_p2'].astype(str)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "Collapsed": "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>hotel_name</th>\n",
       "      <th>review_summary</th>\n",
       "      <th>review_p1</th>\n",
       "      <th>review_p2</th>\n",
       "      <th>score</th>\n",
       "      <th>review</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>Hotel Hilton London Gatwick Airport</td>\n",
       "      <td>Awful...</td>\n",
       "      <td>Its hard to find words that can describe this ...</td>\n",
       "      <td>its just all a bit 'run down'  Poor food, not...</td>\n",
       "      <td>1</td>\n",
       "      <td>Its hard to find words that can describe this ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>7</td>\n",
       "      <td>Hilton London Metropole</td>\n",
       "      <td>Terrible</td>\n",
       "      <td>Absolutely terrible. Cracked ceiling, tiny roo...</td>\n",
       "      <td></td>\n",
       "      <td>1</td>\n",
       "      <td>Absolutely terrible. Cracked ceiling, tiny roo...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>38</td>\n",
       "      <td>Hilton London Euston</td>\n",
       "      <td>Double deposit taken from bank</td>\n",
       "      <td>Very disappointed firstly i checked in online ...</td>\n",
       "      <td></td>\n",
       "      <td>1</td>\n",
       "      <td>Very disappointed firstly i checked in online ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>40</td>\n",
       "      <td>Hilton London Metropole</td>\n",
       "      <td>Super disappointed!</td>\n",
       "      <td>First of all we arrived wanting to park at the...</td>\n",
       "      <td>…</td>\n",
       "      <td>1</td>\n",
       "      <td>First of all we arrived wanting to park at the...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>43</td>\n",
       "      <td>Hilton London Metropole</td>\n",
       "      <td>Dissatisfied Hilton gold number</td>\n",
       "      <td>Hello,  I currently staying @ Hilton Metropole...</td>\n",
       "      <td>…</td>\n",
       "      <td>1</td>\n",
       "      <td>Hello,  I currently staying @ Hilton Metropole...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>48</td>\n",
       "      <td>Hotel Hilton London Gatwick Airport</td>\n",
       "      <td>Wouldn’t refund even though a government ban o...</td>\n",
       "      <td>1.  Hotel gave option to pay on arrival which ...</td>\n",
       "      <td></td>\n",
       "      <td>1</td>\n",
       "      <td>1.  Hotel gave option to pay on arrival which ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>57</td>\n",
       "      <td>Hilton London Euston</td>\n",
       "      <td>Stay away, you can’t trust them</td>\n",
       "      <td>Whether the prices are available on the bar or...</td>\n",
       "      <td></td>\n",
       "      <td>1</td>\n",
       "      <td>Whether the prices are available on the bar or...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>59</td>\n",
       "      <td>Hilton London Euston</td>\n",
       "      <td>Only sty here if nothing else is available in ...</td>\n",
       "      <td>The staff did not seem to be engaged or intere...</td>\n",
       "      <td></td>\n",
       "      <td>1</td>\n",
       "      <td>The staff did not seem to be engaged or intere...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>67</td>\n",
       "      <td>Hilton London Metropole</td>\n",
       "      <td>Rude staff</td>\n",
       "      <td>Restaurant staff were extremely rude. They did...</td>\n",
       "      <td></td>\n",
       "      <td>1</td>\n",
       "      <td>Restaurant staff were extremely rude. They did...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>78</td>\n",
       "      <td>Hilton London Croydon</td>\n",
       "      <td>Just plain awful staff attitude !!!!</td>\n",
       "      <td>Stayed here with work for one night , what a m...</td>\n",
       "      <td>…</td>\n",
       "      <td>1</td>\n",
       "      <td>Stayed here with work for one night , what a m...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                             hotel_name  \\\n",
       "0   Hotel Hilton London Gatwick Airport   \n",
       "7               Hilton London Metropole   \n",
       "38                 Hilton London Euston   \n",
       "40              Hilton London Metropole   \n",
       "43              Hilton London Metropole   \n",
       "48  Hotel Hilton London Gatwick Airport   \n",
       "57                 Hilton London Euston   \n",
       "59                 Hilton London Euston   \n",
       "67              Hilton London Metropole   \n",
       "78                Hilton London Croydon   \n",
       "\n",
       "                                       review_summary  \\\n",
       "0                                            Awful...   \n",
       "7                                           Terrible    \n",
       "38                     Double deposit taken from bank   \n",
       "40                                Super disappointed!   \n",
       "43                    Dissatisfied Hilton gold number   \n",
       "48  Wouldn’t refund even though a government ban o...   \n",
       "57                    Stay away, you can’t trust them   \n",
       "59  Only sty here if nothing else is available in ...   \n",
       "67                                         Rude staff   \n",
       "78               Just plain awful staff attitude !!!!   \n",
       "\n",
       "                                            review_p1  \\\n",
       "0   Its hard to find words that can describe this ...   \n",
       "7   Absolutely terrible. Cracked ceiling, tiny roo...   \n",
       "38  Very disappointed firstly i checked in online ...   \n",
       "40  First of all we arrived wanting to park at the...   \n",
       "43  Hello,  I currently staying @ Hilton Metropole...   \n",
       "48  1.  Hotel gave option to pay on arrival which ...   \n",
       "57  Whether the prices are available on the bar or...   \n",
       "59  The staff did not seem to be engaged or intere...   \n",
       "67  Restaurant staff were extremely rude. They did...   \n",
       "78  Stayed here with work for one night , what a m...   \n",
       "\n",
       "                                            review_p2  score  \\\n",
       "0    its just all a bit 'run down'  Poor food, not...      1   \n",
       "7                                                          1   \n",
       "38                                                         1   \n",
       "40                                                  …      1   \n",
       "43                                                  …      1   \n",
       "48                                                         1   \n",
       "57                                                         1   \n",
       "59                                                         1   \n",
       "67                                                         1   \n",
       "78                                                  …      1   \n",
       "\n",
       "                                               review  \n",
       "0   Its hard to find words that can describe this ...  \n",
       "7   Absolutely terrible. Cracked ceiling, tiny roo...  \n",
       "38  Very disappointed firstly i checked in online ...  \n",
       "40  First of all we arrived wanting to park at the...  \n",
       "43  Hello,  I currently staying @ Hilton Metropole...  \n",
       "48  1.  Hotel gave option to pay on arrival which ...  \n",
       "57  Whether the prices are available on the bar or...  \n",
       "59  The staff did not seem to be engaged or intere...  \n",
       "67  Restaurant staff were extremely rude. They did...  \n",
       "78  Stayed here with work for one night , what a m...  "
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "Collapsed": "false"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "\"Its hard to find words that can describe this experience without being too dismissive so i'll keep it short.  Pros. - Check experience was fine.... The checkin clerk was funny and helpful, and had a lovely manner with the guests. Very efficient. - Location... Very well set for access to Gatwick airport, with a short walk to the terminal for your flights.  Cons - Food...  the hotel not only was packed full and the restaurants were full, but the food itself was incredibly poor and expensive.   We did the buffet... on the evening.... barely anything was truly good tasting..... AVOID - Comfort... Hiltons of late seem to be less and less comfortable, the rooms stark... and without any character. Not the cleanest of rooms either. - its a big hotel.... everywhere is a lengthy walk and its just all a bit 'run down'  Poor food, not the best night sleep, expensive for what it is, tired and in need of significant change.  AVOID!\""
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['review'][0]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "Collapsed": "false"
   },
   "source": [
    "## 1.7 Removing Review Part 1 and Review Part 2 Columns"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "Collapsed": "false"
   },
   "source": [
    "To reduce the number of uneccesary columns, I removed the review part 1 and review part 2 columns."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "Collapsed": "false"
   },
   "outputs": [],
   "source": [
    "df.drop(columns=['review_p1','review_p2'],inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "Collapsed": "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>hotel_name</th>\n",
       "      <th>review_summary</th>\n",
       "      <th>score</th>\n",
       "      <th>review</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>Hotel Hilton London Gatwick Airport</td>\n",
       "      <td>Awful...</td>\n",
       "      <td>1</td>\n",
       "      <td>Its hard to find words that can describe this ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>7</td>\n",
       "      <td>Hilton London Metropole</td>\n",
       "      <td>Terrible</td>\n",
       "      <td>1</td>\n",
       "      <td>Absolutely terrible. Cracked ceiling, tiny roo...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>38</td>\n",
       "      <td>Hilton London Euston</td>\n",
       "      <td>Double deposit taken from bank</td>\n",
       "      <td>1</td>\n",
       "      <td>Very disappointed firstly i checked in online ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>40</td>\n",
       "      <td>Hilton London Metropole</td>\n",
       "      <td>Super disappointed!</td>\n",
       "      <td>1</td>\n",
       "      <td>First of all we arrived wanting to park at the...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>43</td>\n",
       "      <td>Hilton London Metropole</td>\n",
       "      <td>Dissatisfied Hilton gold number</td>\n",
       "      <td>1</td>\n",
       "      <td>Hello,  I currently staying @ Hilton Metropole...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                             hotel_name                   review_summary  \\\n",
       "0   Hotel Hilton London Gatwick Airport                         Awful...   \n",
       "7               Hilton London Metropole                        Terrible    \n",
       "38                 Hilton London Euston   Double deposit taken from bank   \n",
       "40              Hilton London Metropole              Super disappointed!   \n",
       "43              Hilton London Metropole  Dissatisfied Hilton gold number   \n",
       "\n",
       "    score                                             review  \n",
       "0       1  Its hard to find words that can describe this ...  \n",
       "7       1  Absolutely terrible. Cracked ceiling, tiny roo...  \n",
       "38      1  Very disappointed firstly i checked in online ...  \n",
       "40      1  First of all we arrived wanting to park at the...  \n",
       "43      1  Hello,  I currently staying @ Hilton Metropole...  "
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "Collapsed": "false"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(9405, 4)"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "Collapsed": "false"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Int64Index: 9405 entries, 0 to 10293\n",
      "Data columns (total 4 columns):\n",
      "hotel_name        9405 non-null object\n",
      "review_summary    9405 non-null object\n",
      "score             9405 non-null int32\n",
      "review            9405 non-null object\n",
      "dtypes: int32(1), object(3)\n",
      "memory usage: 650.6+ KB\n"
     ]
    }
   ],
   "source": [
    "df.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "Collapsed": "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>score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>count</td>\n",
       "      <td>9405.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>mean</td>\n",
       "      <td>3.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>std</td>\n",
       "      <td>1.414289</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>min</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>25%</td>\n",
       "      <td>2.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>50%</td>\n",
       "      <td>3.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>75%</td>\n",
       "      <td>4.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>max</td>\n",
       "      <td>5.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             score\n",
       "count  9405.000000\n",
       "mean      3.000000\n",
       "std       1.414289\n",
       "min       1.000000\n",
       "25%       2.000000\n",
       "50%       3.000000\n",
       "75%       4.000000\n",
       "max       5.000000"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "Collapsed": "false"
   },
   "source": [
    "## 1.8 Saving Structured Dataset as a CSV"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "Collapsed": "false"
   },
   "source": [
    "I combined the review p1 and review p2 column into one to make future vectorisation much easier, then I saved the cleaned dataset as a csv, for the next stage."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "Collapsed": "false"
   },
   "outputs": [],
   "source": [
    "df.to_csv('Data/2.hotel_reviews_structured.csv',index=False)"
   ]
  }
 ],
 "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": "3.7.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}