amazon-food-review-system.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# [1].LOADING LIBRARIES"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "_cell_guid": "b1076dfc-b9ad-4769-8c92-a6c4dae69d19",
    "_uuid": "8f2839f25d086af736a60e9eeb907d3b93b6e0e5"
   },
   "outputs": [],
   "source": [
    "\n",
    "import warnings\n",
    "warnings.filterwarnings('ignore')\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.model_selection import cross_val_score,train_test_split\n",
    "import nltk\n",
    "#nltk.download('stopwords')\n",
    "from nltk.corpus import stopwords\n",
    "from nltk.stem import PorterStemmer,SnowballStemmer\n",
    "from bs4 import BeautifulSoup\n",
    "from tqdm import tqdm\n",
    "import re\n",
    "import os\n",
    "import sqlite3\n",
    "import time\n",
    "from sklearn.manifold import TSNE\n",
    "import matplotlib.patches as mpatches\n",
    "from sklearn.model_selection import cross_val_score,train_test_split\n",
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "import datetime\n",
    "from sklearn.metrics import confusion_matrix,f1_score,precision_recall_curve,precision_score,recall_score,roc_auc_score,roc_curve,accuracy_score,classification_report\n",
    "\n",
    "import scikitplot.metrics as skplt"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## LOADING DATA & EDA"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of data points in our data (5000, 10)\n"
     ]
    },
    {
     "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>Id</th>\n",
       "      <th>ProductId</th>\n",
       "      <th>UserId</th>\n",
       "      <th>ProfileName</th>\n",
       "      <th>HelpfulnessNumerator</th>\n",
       "      <th>HelpfulnessDenominator</th>\n",
       "      <th>Score</th>\n",
       "      <th>Time</th>\n",
       "      <th>Summary</th>\n",
       "      <th>Text</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>B001E4KFG0</td>\n",
       "      <td>A3SGXH7AUHU8GW</td>\n",
       "      <td>delmartian</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1303862400</td>\n",
       "      <td>Good Quality Dog Food</td>\n",
       "      <td>I have bought several of the Vitality canned d...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>B00813GRG4</td>\n",
       "      <td>A1D87F6ZCVE5NK</td>\n",
       "      <td>dll pa</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1346976000</td>\n",
       "      <td>Not as Advertised</td>\n",
       "      <td>Product arrived labeled as Jumbo Salted Peanut...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>B000LQOCH0</td>\n",
       "      <td>ABXLMWJIXXAIN</td>\n",
       "      <td>Natalia Corres \"Natalia Corres\"</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1219017600</td>\n",
       "      <td>\"Delight\" says it all</td>\n",
       "      <td>This is a confection that has been around a fe...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Id   ProductId          UserId                      ProfileName  \\\n",
       "0   1  B001E4KFG0  A3SGXH7AUHU8GW                       delmartian   \n",
       "1   2  B00813GRG4  A1D87F6ZCVE5NK                           dll pa   \n",
       "2   3  B000LQOCH0   ABXLMWJIXXAIN  Natalia Corres \"Natalia Corres\"   \n",
       "\n",
       "   HelpfulnessNumerator  HelpfulnessDenominator  Score        Time  \\\n",
       "0                     1                       1      1  1303862400   \n",
       "1                     0                       0      0  1346976000   \n",
       "2                     1                       1      1  1219017600   \n",
       "\n",
       "                 Summary                                               Text  \n",
       "0  Good Quality Dog Food  I have bought several of the Vitality canned d...  \n",
       "1      Not as Advertised  Product arrived labeled as Jumbo Salted Peanut...  \n",
       "2  \"Delight\" says it all  This is a confection that has been around a fe...  "
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# loading the data\n",
    "con = sqlite3.connect('database.sqlite')\n",
    "\n",
    "Amazon_Food_Reviews = pd.read_sql_query(\"\"\" SELECT * FROM Reviews WHERE Score != 3 LIMIT 5000\"\"\", con) \n",
    "\n",
    "# Give reviews with Score>3 a positive rating(1), and reviews with a score<3 a negative rating(0).\n",
    "def partition(x):\n",
    "    if x < 3:\n",
    "        return 0\n",
    "    return 1\n",
    "\n",
    "#changing reviews with score less than 3 to be positive and vice-versa\n",
    "actualScore = Amazon_Food_Reviews['Score']\n",
    "positiveNegative = actualScore.map(partition) \n",
    "Amazon_Food_Reviews['Score'] = positiveNegative\n",
    "print(\"Number of data points in our data\", Amazon_Food_Reviews.shape)\n",
    "Amazon_Food_Reviews.head(3)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "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>Id</th>\n",
       "      <th>ProductId</th>\n",
       "      <th>UserId</th>\n",
       "      <th>ProfileName</th>\n",
       "      <th>HelpfulnessNumerator</th>\n",
       "      <th>HelpfulnessDenominator</th>\n",
       "      <th>Score</th>\n",
       "      <th>Time</th>\n",
       "      <th>Summary</th>\n",
       "      <th>Text</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>B001E4KFG0</td>\n",
       "      <td>A3SGXH7AUHU8GW</td>\n",
       "      <td>delmartian</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1303862400</td>\n",
       "      <td>Good Quality Dog Food</td>\n",
       "      <td>I have bought several of the Vitality canned d...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>B00813GRG4</td>\n",
       "      <td>A1D87F6ZCVE5NK</td>\n",
       "      <td>dll pa</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1346976000</td>\n",
       "      <td>Not as Advertised</td>\n",
       "      <td>Product arrived labeled as Jumbo Salted Peanut...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>B000LQOCH0</td>\n",
       "      <td>ABXLMWJIXXAIN</td>\n",
       "      <td>Natalia Corres \"Natalia Corres\"</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1219017600</td>\n",
       "      <td>\"Delight\" says it all</td>\n",
       "      <td>This is a confection that has been around a fe...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>B000UA0QIQ</td>\n",
       "      <td>A395BORC6FGVXV</td>\n",
       "      <td>Karl</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>1307923200</td>\n",
       "      <td>Cough Medicine</td>\n",
       "      <td>If you are looking for the secret ingredient i...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>B006K2ZZ7K</td>\n",
       "      <td>A1UQRSCLF8GW1T</td>\n",
       "      <td>Michael D. Bigham \"M. Wassir\"</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1350777600</td>\n",
       "      <td>Great taffy</td>\n",
       "      <td>Great taffy at a great price.  There was a wid...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Id   ProductId          UserId                      ProfileName  \\\n",
       "0   1  B001E4KFG0  A3SGXH7AUHU8GW                       delmartian   \n",
       "1   2  B00813GRG4  A1D87F6ZCVE5NK                           dll pa   \n",
       "2   3  B000LQOCH0   ABXLMWJIXXAIN  Natalia Corres \"Natalia Corres\"   \n",
       "3   4  B000UA0QIQ  A395BORC6FGVXV                             Karl   \n",
       "4   5  B006K2ZZ7K  A1UQRSCLF8GW1T    Michael D. Bigham \"M. Wassir\"   \n",
       "\n",
       "   HelpfulnessNumerator  HelpfulnessDenominator  Score        Time  \\\n",
       "0                     1                       1      1  1303862400   \n",
       "1                     0                       0      0  1346976000   \n",
       "2                     1                       1      1  1219017600   \n",
       "3                     3                       3      0  1307923200   \n",
       "4                     0                       0      1  1350777600   \n",
       "\n",
       "                 Summary                                               Text  \n",
       "0  Good Quality Dog Food  I have bought several of the Vitality canned d...  \n",
       "1      Not as Advertised  Product arrived labeled as Jumbo Salted Peanut...  \n",
       "2  \"Delight\" says it all  This is a confection that has been around a fe...  \n",
       "3         Cough Medicine  If you are looking for the secret ingredient i...  \n",
       "4            Great taffy  Great taffy at a great price.  There was a wid...  "
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Amazon_Food_Reviews.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "000.280 % of data removed during duplicates removal operation\n"
     ]
    }
   ],
   "source": [
    "Amazon_Food_Reviews_Sorted_Data= Amazon_Food_Reviews.sort_values('ProductId',axis=0, inplace=False, kind='quicksort', na_position='last')\n",
    "Amazon_Food_Reviews_Duplicates_Dropped= Amazon_Food_Reviews_Sorted_Data.drop_duplicates(subset={\"UserId\",\"ProfileName\",\"Time\",\"Text\"},keep='first',inplace=False)\n",
    "print(\"{:07.3f} % of data removed during duplicates removal operation\".format((1-(Amazon_Food_Reviews_Duplicates_Dropped.shape[0])/(Amazon_Food_Reviews_Sorted_Data.shape[0]))*100))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "000.000 % of data removed\n"
     ]
    }
   ],
   "source": [
    "Amazon_Food_Reviews_Preprocessed=Amazon_Food_Reviews_Duplicates_Dropped[Amazon_Food_Reviews_Duplicates_Dropped.HelpfulnessNumerator<=Amazon_Food_Reviews_Duplicates_Dropped.HelpfulnessDenominator]\n",
    "print(\"{:07.3f} % of data removed\".format(np.round((1-(Amazon_Food_Reviews_Preprocessed.shape[0])/(Amazon_Food_Reviews_Duplicates_Dropped.shape[0]))*100)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# [3].  Text Preprocessing.\n",
    "\n",
    "Now that we have finished deduplication our data requires some preprocessing before we go on further with analysis and making the prediction model.\n",
    "\n",
    "Hence in the Preprocessing phase we do the following in the order below:-\n",
    "\n",
    "1. Begin by removing the html tags\n",
    "2. Remove any punctuations or limited set of special characters like , or . or # etc.\n",
    "3. Check if the word is made up of english letters and is not alpha-numeric\n",
    "4. Check to see if the length of the word is greater than 2 (as it was researched that there is no adjective in 2-letters)\n",
    "5. Convert the word to lowercase\n",
    "6. Remove Stopwords\n",
    "7. Finally Snowball Stemming the word (it was obsereved to be better than Porter Stemming)<br>\n",
    "\n",
    "After which we collect the words used to describe positive and negative reviews"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "def DeContraction(document):\n",
    "    document= re.sub(r\"won\\'t\",\"will not\",document)\n",
    "    document= re.sub(r\"can\\'t\",\"can not\",document)\n",
    "    document= re.sub(r\"\\'t\",\"not\",document)\n",
    "    document= re.sub(r\"\\'m\",\"am\",document)\n",
    "    document= re.sub(r\"\\'re\",\"are\",document)\n",
    "    document= re.sub(r\"\\'ve\",\"have\",document)\n",
    "    document= re.sub(r\"\\'d\",\"would\",document)\n",
    "    document= re.sub(r\"\\'s\",\"is\",document)\n",
    "    document= re.sub(r\"\\'ll\",\"will\",document)\n",
    "    return document"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "Custom_StopWords = [word for word in stopwords.words('english') if word not in ['no','nor','not']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "from tqdm import tqdm\n",
    "def Preprocessing(Text_Data):\n",
    "    Preprocessed_Text =[]\n",
    "    for document in tqdm(Text_Data):\n",
    "        document= re.sub(r\"http\\S+\",\"\",document)\n",
    "        document = BeautifulSoup(document,'html').get_text()\n",
    "        document=DeContraction(document)\n",
    "        document = re.sub(r\"\\S*\\d\\S*\",\"\",document).strip()\n",
    "        document = re.sub(\"[^a-zA-Z]+\",' ',document)\n",
    "        document = ' '.join(word.lower() for word in document.split() if word.lower() not in Custom_StopWords)\n",
    "        Preprocessed_Text.append(document)\n",
    "    return Preprocessed_Text"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|█████████████████████████████████████████████████████████████████████████████| 4986/4986 [00:12<00:00, 397.08it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "860 :ihave tried mixes pancakes theyhave awful one really good buy stock co people buying product iam sure\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "Preprocessed_Amazon_Reviews = Preprocessing(Amazon_Food_Reviews_Preprocessed['Text'].values)\n",
    "np.random.seed(42)\n",
    "print(\"{} :{}\".format(np.random.randint(Amazon_Food_Reviews_Preprocessed.shape[0]),\n",
    "                      Preprocessed_Amazon_Reviews[np.random.randint(Amazon_Food_Reviews_Preprocessed.shape[0])]))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# [4] FEATURIZATION OF ENTIRE DATASET"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## [4.1] BAG OF WORDS"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Type of Output:<class 'scipy.sparse.csr.csr_matrix'>,Size:(4986, 13278)\n"
     ]
    }
   ],
   "source": [
    "from sklearn.feature_extraction.text import CountVectorizer\n",
    "cnt_vectorizer= CountVectorizer()\n",
    "Count_Vectorized_Reviews=cnt_vectorizer.fit_transform(Preprocessed_Amazon_Reviews)\n",
    "print(\"Type of Output:{},Size:{}\".format(type(Count_Vectorized_Reviews),Count_Vectorized_Reviews.get_shape()))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## [4.2] Bi-Grams and n-Grams."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Type of Output:<class 'scipy.sparse.csr.csr_matrix'>,Size:(4986, 3156)\n"
     ]
    }
   ],
   "source": [
    "#Count Vector With Bi-Grams,Words considering whose frequency greater than 10\n",
    "Cnt_Vectorizer_Bigrams=CountVectorizer(ngram_range=(1,2),min_df=10)\n",
    "Count_Vectorized_Reviews_Bigram = Cnt_Vectorizer_Bigrams.fit_transform(Preprocessed_Amazon_Reviews)\n",
    "print(\"Type of Output:{},Size:{}\".format(type(Count_Vectorized_Reviews_Bigram),Count_Vectorized_Reviews_Bigram.get_shape()))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## [4.3] TF-IDF"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Type of Output:<class 'scipy.sparse.csr.csr_matrix'>,Size:(4986, 3156)\n"
     ]
    }
   ],
   "source": [
    "from sklearn.feature_extraction.text import TfidfVectorizer\n",
    "tfidf_vectorizer_bigram = TfidfVectorizer(ngram_range=(1,2),min_df=10)\n",
    "TFIDF_Vectorized_Reviews_Bigram=tfidf_vectorizer_bigram.fit_transform(Preprocessed_Amazon_Reviews)\n",
    "print(\"Type of Output:{},Size:{}\".format(type(TFIDF_Vectorized_Reviews_Bigram),TFIDF_Vectorized_Reviews_Bigram.get_shape()))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## [4.4] Word2Vec"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "List_Of_document=[]\n",
    "for document in Preprocessed_Amazon_Reviews:\n",
    "    List_Of_document.append(document.split())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "from gensim.models import Word2Vec\n",
    "def Train_Word2Vec(List_Of_document,want_to_train_own_word2vec_model=True):\n",
    "    if want_to_train_own_word2vec_model:\n",
    "        return Word2Vec(List_Of_document,min_count=5,size=128,workers=os.cpu_count()-2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "Word2Vec_Model = Train_Word2Vec(List_Of_document)\n",
    "w2v_words=list(Word2Vec_Model.wv.vocab)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### [4.4.1] AVG WORD2VEC"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "def Weighted_Word2Vec(Word2Vec_Model,List_Of_document):\n",
    "    Documents_Vector=[]\n",
    "    for document in tqdm(List_Of_document):\n",
    "        Document_Vector=np.zeros(128)\n",
    "        count_words=0\n",
    "        for word in document:\n",
    "            if word in w2v_words:\n",
    "                word_vector=Word2Vec_Model.wv[word]\n",
    "                Document_Vector+=word_vector\n",
    "                count_words+=1\n",
    "        if count_words!=0:\n",
    "            Document_Vector/=count_words\n",
    "        Documents_Vector.append(Document_Vector)\n",
    "    return Documents_Vector"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|█████████████████████████████████████████████████████████████████████████████| 4986/4986 [00:07<00:00, 695.68it/s]\n"
     ]
    }
   ],
   "source": [
    "Weighted_Word2Vec_Vectorization=Weighted_Word2Vec(Word2Vec_Model,List_Of_document)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### [4.4.2] TFIDF WORD2VEC"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "tfidf_model =TfidfVectorizer()\n",
    "tfidf_model.fit(Preprocessed_Amazon_Reviews)\n",
    "IDF_Dictionary= dict(zip(tfidf_model.get_feature_names(),tfidf_model.idf_))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "def TFIDF_WORD2VEC(tfidf_model,IDF_Dictionary,List_Of_document):\n",
    "    TFIDF_FEATURES = tfidf_model.get_feature_names()\n",
    "    TFIDF_Documents_Vector=[]\n",
    "    row=0\n",
    "    for Document in tqdm(List_Of_document):\n",
    "        Document_Vector=np.zeros(128)\n",
    "        weighted_sum=0\n",
    "        for word in Document:\n",
    "            if word in w2v_words and word in TFIDF_FEATURES:\n",
    "                word_vector=Word2Vec_Model.wv[word]\n",
    "                word_tfidf= IDF_Dictionary[word]*(Document.count(word)/len(Document))\n",
    "                Document_Vector+=word_vector*word_tfidf\n",
    "                weighted_sum+=word_tfidf\n",
    "        if weighted_sum!=0:\n",
    "            Document_Vector/=weighted_sum\n",
    "        TFIDF_Documents_Vector.append(Document_Vector)\n",
    "        row+=1\n",
    "    return TFIDF_Documents_Vector"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|█████████████████████████████████████████████████████████████████████████████| 4986/4986 [00:46<00:00, 106.34it/s]\n"
     ]
    }
   ],
   "source": [
    "TFIDF_Word2Vec_Vectorization=TFIDF_WORD2VEC(tfidf_model,IDF_Dictionary,List_Of_document)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# [ 5] VISUALIZATION"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "import time\n",
    "os.mkdir('Visualization/')\n",
    "def Visualization(Dataset,Score,Title,no_of_dims=2,perplexity=50,learning_rate=200):\n",
    "    plt.figure(figsize=(12,10))\n",
    "    TSNE_Model = TSNE(n_components=no_of_dims,perplexity=perplexity,learning_rate=learning_rate,n_jobs=-1,n_iter=5000)\n",
    "    start_time=time.time()\n",
    "    X_TSNE=TSNE_Model.fit_transform(Dataset)\n",
    "    stop_time=time.time()\n",
    "    Data = np.hstack((X_TSNE,Score.reshape(-1,1)))\n",
    "    TSNE_DF=pd.DataFrame(Data,columns=['Dimension_x','Dimension_y','Score'])\n",
    "    colors = {0:'red', 1:'blue'}\n",
    "    plt.scatter(TSNE_DF['Dimension_x'], TSNE_DF['Dimension_y'], c=TSNE_DF['Score'].apply(lambda x: colors[x]))\n",
    "    plt.title(Title)\n",
    "    green_patch = mpatches.Patch(color='green', label='perplexity :'+str(perplexity))\n",
    "    black_patch = mpatches.Patch(color='black', label='learning rate :'+str(learning_rate))\n",
    "    plt.xlabel('Dimension_x')\n",
    "    plt.ylabel('Dimension_y')\n",
    "    plt.legend(handles=[green_patch, black_patch])\n",
    "    Image = os.path.join('Visualization/',Title+'n_components'+'_'+str(no_of_dims)+'_'+'perplexity'+'_'+str(perplexity)+'_'+'learning_rate'+'_'+str(learning_rate)+'.png')\n",
    "    plt.savefig(fname =Image,dpi=500,format='png')\n",
    "    Total_Time=(stop_time-start_time)/60\n",
    "    print('Execution Time:{}'.format(Total_Time))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## [5.1] COUNT VECTORIZATION"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Execution Time:9.093512686093648\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "Visualization(Dataset=Count_Vectorized_Reviews.toarray(),Score=Amazon_Food_Reviews_Preprocessed['Score'].values,Title='Count Vectoization TSNE',no_of_dims=2,perplexity=10,learning_rate=200)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Execution Time:7.80785372654597\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "Visualization(Dataset=Count_Vectorized_Reviews.toarray(),Score=Amazon_Food_Reviews_Preprocessed['Score'].values,Title='Count Vectoization TSNE',no_of_dims=2,perplexity=30,learning_rate=200)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "Visualization(Dataset=Count_Vectorized_Reviews.toarray(),Score=Amazon_Food_Reviews_Preprocessed['Score'].values,Title='Count Vectoization TSNE',no_of_dims=2,perplexity=20,learning_rate=200)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "Visualization(Dataset=Count_Vectorized_Reviews.toarray(),Score=Amazon_Food_Reviews_Preprocessed['Score'].values,Title='Count Vectoization TSNE',no_of_dims=2,perplexity=40,learning_rate=200)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "Visualization(Dataset=Count_Vectorized_Reviews.toarray(),Score=Amazon_Food_Reviews_Preprocessed['Score'].values,Title='Count Vectoization TSNE')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "Visualization(Dataset=Count_Vectorized_Reviews.toarray(),Score=Amazon_Food_Reviews_Preprocessed['Score'].values,Title='Count Vectoization TSNE',no_of_dims=2,perplexity=100,learning_rate=200)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## [5.2] BIGRAM COUNT VECTORIZATION "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "Visualization(Dataset=Count_Vectorized_Reviews_Bigram.toarray(),Score=Amazon_Food_Reviews_Preprocessed['Score'].values,Title='BiGRAM Count Vectoization TSNE',no_of_dims=2,perplexity=10,learning_rate=200)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "Visualization(Dataset=Count_Vectorized_Reviews_Bigram.toarray(),Score=Amazon_Food_Reviews_Preprocessed['Score'].values,Title='BiGRAM Count Vectoization TSNE',no_of_dims=2,perplexity=20,learning_rate=200)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "Visualization(Dataset=Count_Vectorized_Reviews_Bigram.toarray(),Score=Amazon_Food_Reviews_Preprocessed['Score'].values,Title='BiGRAM Count Vectoization TSNE',no_of_dims=2,perplexity=30,learning_rate=200)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "Visualization(Dataset=Count_Vectorized_Reviews_Bigram.toarray(),Score=Amazon_Food_Reviews_Preprocessed['Score'].values,Title='BiGRAM Count Vectoization TSNE',no_of_dims=2,perplexity=40,learning_rate=200)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "Visualization(Dataset=Count_Vectorized_Reviews_Bigram.toarray(),Score=Amazon_Food_Reviews_Preprocessed['Score'].values,Title='BiGRAM Count Vectoization TSNE')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "Visualization(Dataset=Count_Vectorized_Reviews_Bigram.toarray(),Score=Amazon_Food_Reviews_Preprocessed['Score'].values,Title='BiGRAM Count Vectoization TSNE',no_of_dims=2,perplexity=100,learning_rate=200)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## [5.3] TFIDF BIGRAM VECTORIZATION "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "Visualization(Dataset=TFIDF_Vectorized_Reviews_Bigram.toarray(),Score=Amazon_Food_Reviews_Preprocessed['Score'].values,Title='BiGRAM TFIDF Vectoization TSNE',no_of_dims=2,perplexity=10,learning_rate=200)"
   ]
  },
  {
   "cell_type": "raw",
   "metadata": {},
   "source": [
    "Visualization(Dataset=TFIDF_Vectorized_Reviews_Bigram.toarray(),Score=Amazon_Food_Reviews_Preprocessed['Score'].values,Title='BiGRAM TFIDF Vectoization TSNE',no_of_dims=2,perplexity=20,learning_rate=200)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "Visualization(Dataset=TFIDF_Vectorized_Reviews_Bigram.toarray(),Score=Amazon_Food_Reviews_Preprocessed['Score'].values,Title='BiGRAM TFIDF Vectoization TSNE',no_of_dims=2,perplexity=30,learning_rate=200)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "Visualization(Dataset=TFIDF_Vectorized_Reviews_Bigram.toarray(),Score=Amazon_Food_Reviews_Preprocessed['Score'].values,Title='BiGRAM TFIDF Vectoization TSNE',no_of_dims=2,perplexity=40,learning_rate=200)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "Visualization(Dataset=TFIDF_Vectorized_Reviews_Bigram.toarray(),Score=Amazon_Food_Reviews_Preprocessed['Score'].values,Title='BiGRAM TFIDF Vectoization TSNE')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "Visualization(Dataset=TFIDF_Vectorized_Reviews_Bigram.toarray(),Score=Amazon_Food_Reviews_Preprocessed['Score'].values,Title='BiGRAM TFIDF Vectoization TSNE',no_of_dims=2,perplexity=100,learning_rate=200)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## [5.4.1] AVG WEIGHTED WORD2VEC VECTORIZATION "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "Visualization(Dataset=Weighted_Word2Vec_Vectorization,Score=Amazon_Food_Reviews_Preprocessed['Score'].values,Title='AVG WORD2VEC Vectoization TSNE',no_of_dims=2,perplexity=10,learning_rate=200)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "Visualization(Dataset=Weighted_Word2Vec_Vectorization,Score=Amazon_Food_Reviews_Preprocessed['Score'].values,Title='AVG WORD2VEC Vectoization TSNE',no_of_dims=2,perplexity=20,learning_rate=200)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "Visualization(Dataset=Weighted_Word2Vec_Vectorization,Score=Amazon_Food_Reviews_Preprocessed['Score'].values,Title='AVG WORD2VEC Vectoization TSNE',no_of_dims=2,perplexity=30,learning_rate=200)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "Visualization(Dataset=Weighted_Word2Vec_Vectorization,Score=Amazon_Food_Reviews_Preprocessed['Score'].values,Title='AVG WORD2VEC Vectoization TSNE',no_of_dims=2,perplexity=40,learning_rate=200)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "Visualization(Dataset=Weighted_Word2Vec_Vectorization,Score=Amazon_Food_Reviews_Preprocessed['Score'].values,Title='AVG WORD2VEC Vectoization TSNE')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "Visualization(Dataset=Weighted_Word2Vec_Vectorization,Score=Amazon_Food_Reviews_Preprocessed['Score'].values,Title='AVG WORD2VEC Vectoization TSNE',no_of_dims=2,perplexity=100,learning_rate=200)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## [5.4.2] TFIDF WEIGHTED WORD2VEC VECTORIZATION "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "Visualization(Dataset=TFIDF_Word2Vec_Vectorization,Score=Amazon_Food_Reviews_Preprocessed['Score'].values,Title='TFIDF WORD2VEC Vectoization TSNE',no_of_dims=2,perplexity=10,learning_rate=200)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "Visualization(Dataset=TFIDF_Word2Vec_Vectorization,Score=Amazon_Food_Reviews_Preprocessed['Score'].values,Title='TFIDF WORD2VEC Vectoization TSNE',no_of_dims=2,perplexity=20,learning_rate=200)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "Visualization(Dataset=TFIDF_Word2Vec_Vectorization,Score=Amazon_Food_Reviews_Preprocessed['Score'].values,Title='TFIDF WORD2VEC Vectoization TSNE',no_of_dims=2,perplexity=30,learning_rate=200)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "Visualization(Dataset=TFIDF_Word2Vec_Vectorization,Score=Amazon_Food_Reviews_Preprocessed['Score'].values,Title='TFIDF WORD2VEC Vectoization TSNE',no_of_dims=2,perplexity=40,learning_rate=200)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "Visualization(Dataset=TFIDF_Word2Vec_Vectorization,Score=Amazon_Food_Reviews_Preprocessed['Score'].values,Title='TFIDF WORD2VEC Vectoization TSNE')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "Visualization(Dataset=TFIDF_Word2Vec_Vectorization,Score=Amazon_Food_Reviews_Preprocessed['Score'].values,Title='TFIDF WORD2VEC Vectoization TSNE',no_of_dims=2,perplexity=100,learning_rate=200)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b>Observation:-</b> This plot also looks like the bow, tfidf and avg word2vec.Both +ve and -ve reviwes are not well seperated they overlapped each other."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b> Conclusions:- </b>\n",
    "\n",
    "1. AS none of TSNE representation gives a well separated both +ve and -ve reviews.\n",
    "2. We can not simply draw a plane to separate -ve and +ve reviews. Although, By looking at only visual representation of data \n",
    "    we can not take decision whether to draw a plane or not.\n",
    "3. We will have some alternative method by that we will look at into this problem like how we can separate -ve and +ve reviews."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# [6] MODELLING"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## [6.1] K-NEAREST NEIGHBOURS"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## TIME BASED SPLITTING"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "def Cleaning_Pipeline(filename,FullDataLoad=False):\n",
    "    Amazon_Food_Reviews,Table_Connection = LoadData(Dataset_Name,FullDataLoad=FullDataLoad)\n",
    "    Amazon_Food_Reviews['Score']= Amazon_Food_Reviews['Score'].apply(lambda x: 1 if x>3 else 0)\n",
    "    print(Amazon_Food_Reviews.head())\n",
    "    \n",
    "    Amazon_Food_Reviews_Sorted_Data= Amazon_Food_Reviews.sort_values('ProductId',axis=0, inplace=False, kind='quicksort', na_position='last')\n",
    "    Amazon_Food_Reviews_Duplicates_Dropped= Amazon_Food_Reviews_Sorted_Data.drop_duplicates(subset={\"UserId\",\"ProfileName\",\"Time\",\"Text\"},keep='first',inplace=False)\n",
    "    print(\"{:07.3f} % of data removed during duplicates removal operation\".format((1-(Amazon_Food_Reviews_Duplicates_Dropped.shape[0])/(Amazon_Food_Reviews_Sorted_Data.shape[0]))*100))\n",
    "    Amazon_Food_Reviews_Preprocessed=Amazon_Food_Reviews_Duplicates_Dropped[Amazon_Food_Reviews_Duplicates_Dropped.HelpfulnessNumerator<=Amazon_Food_Reviews_Duplicates_Dropped.HelpfulnessDenominator]\n",
    "    print(\"{:07.3f} % of data removed\".format(np.round((1-(Amazon_Food_Reviews_Preprocessed.shape[0])/(Amazon_Food_Reviews_Duplicates_Dropped.shape[0]))*100)))\n",
    "    return Amazon_Food_Reviews_Preprocessed"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "Dataset_Name='database.sqlite'\n",
    "#cleaning 5000 records\n",
    "Amazon_Food_Reviews_Preprocessed= Cleaning_Pipeline(Dataset_Name,FullDataLoad=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "def Balanced_Data(Total_Dataset):\n",
    "    return pd.concat([Total_Dataset[Total_Dataset['Score']==1].sample(frac=0.035,random_state=1),Total_Dataset[Total_Dataset['Score']==0].sample(frac=0.15,random_state=1)],axis=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "Balanced_Amazon_Data_Preprocessed=Balanced_Data(Total_Dataset=Amazon_Food_Reviews_Preprocessed)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "Balanced_Data_Time_Sorted=Balanced_Amazon_Data_Preprocessed.sort_values('Time',axis=0,kind='quicksort',ascending=True).reset_index(drop=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "Balanced_Data_Time_Sorted['Score'].value_counts().plot(kind='bar')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "def Find_Optimal_K_In_KNN(X_Train,Y_Train,K_Values_List):\n",
    "    #creating odd list of K values\n",
    "    neighbors= list(filter(lambda k: k%2!=0,K_Values_List))\n",
    "    CV_Scores=[]\n",
    "    for k in tqdm(neighbors):\n",
    "        knn = KNeighborsClassifier(n_neighbors=k,n_jobs=-1)\n",
    "        scores = cross_val_score(knn,X_Train,Y_Train,cv=10,scoring='accuracy',n_jobs=-1)\n",
    "        CV_Scores.append(scores.mean())\n",
    "    MSE = [1-x for x in CV_Scores]\n",
    "    Optimal_K = neighbors[MSE.index(min(MSE))]\n",
    "    print('\\nThe optimal number of neighbors is {}'.format(Optimal_K))\n",
    "    plt.figure(figsize=(12,10))\n",
    "    plt.plot(neighbors,MSE,color='blue', linestyle='dashed',marker='o',markerfacecolor='red', markersize=10)\n",
    "    plt.title('Error Rate vs. K Value')\n",
    "    plt.xlabel('K')\n",
    "    plt.ylabel('Error Rate')\n",
    "    print(\"the misclassification error for each k value is : \", np.round(MSE,3))\n",
    "    return Optimal_K"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## [6.1.1] COUNT VECTORIZATION"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_Train,X_Test,Y_Train,Y_Test = train_test_split(Balanced_Data_Time_Sorted['Text'],Balanced_Data_Time_Sorted['Score'],test_size=0.25)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.preprocessing import StandardScaler\n",
    "\n",
    "cnt_vect = CountVectorizer(ngram_range=(1,2),min_df=10,max_features=5000)\n",
    "X_Train_Features=cnt_vect.fit_transform(X_Train)\n",
    "X_Test_Features = cnt_vect.transform(X_Test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "K_Neighbours_List = list(range(0,20))\n",
    "Optimal_K = Find_Optimal_K_In_KNN(X_Train_Features.toarray() ,Y_Train,K_Neighbours_List)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "knn = KNeighborsClassifier(n_neighbors=Optimal_K)\n",
    "knn.fit(X_Train_Features, Y_Train)\n",
    "kn_pred = knn.predict(X_Test_Features)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "knn_confusion_matrix=confusion_matrix(Y_Test,kn_pred)\n",
    "skplt.plot_confusion_matrix(Y_Test,kn_pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.metrics import accuracy_score,classification_report\n",
    "print(accuracy_score(Y_Test,kn_pred))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "print(classification_report(Y_Test,kn_pred))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## [6.1.1] TFIDF VECTORIZATION"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.feature_extraction.text import TfidfVectorizer\n",
    "#X_Train,X_Test,Y_Train,Y_Test = train_test_split(Balanced_Data_Time_Sorted['Text'],Balanced_Data_Time_Sorted['Score'],test_size=0.25)\n",
    "TFIDF_vect = TfidfVectorizer(ngram_range=(1,2),min_df=10,max_features=5000)\n",
    "X_Train_Features=TFIDF_vect.fit_transform(X_Train)\n",
    "X_Test_Features = TFIDF_vect.transform(X_Test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "K_Neighbours_List = list(range(0,20))\n",
    "Optimal_K = Find_Optimal_K_In_KNN(X_Train_Features.toarray() ,Y_Train,K_Neighbours_List)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "knn = KNeighborsClassifier(n_neighbors=Optimal_K)\n",
    "knn.fit(X_Train_Features, Y_Train)\n",
    "knn_TFIDF_pred = knn.predict(X_Test_Features)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "knn_confusion_matrix=confusion_matrix(Y_Test,knn_TFIDF_pred)\n",
    "skplt.plot_confusion_matrix(Y_Test,knn_TFIDF_pred)\n",
    "print(accuracy_score(Y_Test,knn_TFIDF_pred))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "print(classification_report(Y_Test,knn_TFIDF_pred))"
   ]
  },
  {
   "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": "3.6.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}