recommender_system.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Import Libraries"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import num2words\n",
    "import re\n",
    "import contractions\n",
    "import string\n",
    "\n",
    "import nltk\n",
    "from nltk.tokenize import sent_tokenize, word_tokenize, TweetTokenizer\n",
    "from nltk.corpus import words, stopwords\n",
    "from nltk.metrics.distance import jaccard_distance, edit_distance\n",
    "from nltk.util import ngrams\n",
    "from nltk.stem import WordNetLemmatizer\n",
    "\n",
    "from sklearn.metrics.pairwise import *\n",
    "from sklearn.feature_extraction.text import TfidfVectorizer, CountVectorizer\n",
    "from sklearn.model_selection import train_test_split, GridSearchCV\n",
    "\n",
    "from flair.data import *\n",
    "from flair.nn import *\n",
    "from flair.embeddings import *\n",
    "tagger = Classifier.load('sentiment-fast')\n",
    "glove_embeddings = WordEmbeddings('glove')\n",
    "\n",
    "import gensim\n",
    "from gensim.models.fasttext import FastText\n",
    "from gensim.models import Word2Vec"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Loading Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Index(['name', 'colour', 'brand', 'description', 'Body or Garment Size',\n",
      "       'Neck', 'Occasion', 'Print or Pattern Type', 'Sleeve Length',\n",
      "       'Sustainable', 'Wash Care', 'Weave Type', 'Fabric', 'Hemline', 'Length',\n",
      "       'Pattern', 'Center Front Open', 'Closure', 'Surface Styling', 'Type',\n",
      "       'Sport'],\n",
      "      dtype='object')\n",
      "name                        0\n",
      "colour                      0\n",
      "brand                       0\n",
      "description                 0\n",
      "Body or Garment Size        0\n",
      "Neck                      179\n",
      "Occasion                    0\n",
      "Print or Pattern Type       8\n",
      "Sleeve Length               0\n",
      "Sustainable                 0\n",
      "Wash Care                   0\n",
      "Weave Type                  0\n",
      "Fabric                      0\n",
      "Hemline                     0\n",
      "Length                      0\n",
      "Pattern                     0\n",
      "Center Front Open           0\n",
      "Closure                  2091\n",
      "Surface Styling          4414\n",
      "Type                      554\n",
      "Sport                    1818\n",
      "dtype: int64\n",
      "(14309, 21)\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>name</th>\n",
       "      <th>colour</th>\n",
       "      <th>brand</th>\n",
       "      <th>description</th>\n",
       "      <th>Body or Garment Size</th>\n",
       "      <th>Neck</th>\n",
       "      <th>Occasion</th>\n",
       "      <th>Print or Pattern Type</th>\n",
       "      <th>Sleeve Length</th>\n",
       "      <th>Sustainable</th>\n",
       "      <th>...</th>\n",
       "      <th>Weave Type</th>\n",
       "      <th>Fabric</th>\n",
       "      <th>Hemline</th>\n",
       "      <th>Length</th>\n",
       "      <th>Pattern</th>\n",
       "      <th>Center Front Open</th>\n",
       "      <th>Closure</th>\n",
       "      <th>Surface Styling</th>\n",
       "      <th>Type</th>\n",
       "      <th>Sport</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Khushal K Women Black Ethnic Motifs Printed Ku...</td>\n",
       "      <td>Black</td>\n",
       "      <td>Khushal K</td>\n",
       "      <td>Black printed Kurta with Palazzos with dupatta...</td>\n",
       "      <td>Garment Measurements in</td>\n",
       "      <td>Mandarin Collar</td>\n",
       "      <td>Festive</td>\n",
       "      <td>Ethnic Motifs</td>\n",
       "      <td>Three-Quarter Sleeves</td>\n",
       "      <td>Regular</td>\n",
       "      <td>...</td>\n",
       "      <td>Machine Weave</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>InWeave Women Orange Solid Kurta with Palazzos...</td>\n",
       "      <td>Orange</td>\n",
       "      <td>InWeave</td>\n",
       "      <td>Orange solid Kurta with Palazzos with dupattaK...</td>\n",
       "      <td>Garment Measurements in</td>\n",
       "      <td>Square Neck</td>\n",
       "      <td>Fusion</td>\n",
       "      <td>Solid</td>\n",
       "      <td>Sleeveless</td>\n",
       "      <td>Regular</td>\n",
       "      <td>...</td>\n",
       "      <td>Machine Weave</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>Anubhutee Women Navy Blue Ethnic Motifs Embroi...</td>\n",
       "      <td>Navy Blue</td>\n",
       "      <td>Anubhutee</td>\n",
       "      <td>Navy blue embroidered Kurta with Trousers with...</td>\n",
       "      <td>Garment Measurements in</td>\n",
       "      <td>Round Neck</td>\n",
       "      <td>Daily</td>\n",
       "      <td>Ethnic Motifs</td>\n",
       "      <td>Three-Quarter Sleeves</td>\n",
       "      <td>Regular</td>\n",
       "      <td>...</td>\n",
       "      <td>Machine Weave</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>Nayo Women Red Floral Printed Kurta With Trous...</td>\n",
       "      <td>Red</td>\n",
       "      <td>Nayo</td>\n",
       "      <td>Red printed kurta with trouser and dupattaKurt...</td>\n",
       "      <td>Garment Measurements in</td>\n",
       "      <td>Round Neck</td>\n",
       "      <td>Daily</td>\n",
       "      <td>Ethnic Motifs</td>\n",
       "      <td>Three-Quarter Sleeves</td>\n",
       "      <td>Regular</td>\n",
       "      <td>...</td>\n",
       "      <td>Machine Weave</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>AHIKA Women Black &amp; Green Printed Straight Kurta</td>\n",
       "      <td>Black</td>\n",
       "      <td>AHIKA</td>\n",
       "      <td>Black and green printed straight kurta, has a ...</td>\n",
       "      <td>Garment Measurements in</td>\n",
       "      <td>Round Neck</td>\n",
       "      <td>Daily</td>\n",
       "      <td>Ethnic Motifs</td>\n",
       "      <td>Three-Quarter Sleeves</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>Machine Weave</td>\n",
       "      <td>Cotton</td>\n",
       "      <td>Straight</td>\n",
       "      <td>Calf Length</td>\n",
       "      <td>Printed</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 × 21 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                name     colour      brand  \\\n",
       "0  Khushal K Women Black Ethnic Motifs Printed Ku...      Black  Khushal K   \n",
       "1  InWeave Women Orange Solid Kurta with Palazzos...     Orange    InWeave   \n",
       "2  Anubhutee Women Navy Blue Ethnic Motifs Embroi...  Navy Blue  Anubhutee   \n",
       "3  Nayo Women Red Floral Printed Kurta With Trous...        Red       Nayo   \n",
       "4   AHIKA Women Black & Green Printed Straight Kurta      Black      AHIKA   \n",
       "\n",
       "                                         description     Body or Garment Size  \\\n",
       "0  Black printed Kurta with Palazzos with dupatta...  Garment Measurements in   \n",
       "1  Orange solid Kurta with Palazzos with dupattaK...  Garment Measurements in   \n",
       "2  Navy blue embroidered Kurta with Trousers with...  Garment Measurements in   \n",
       "3  Red printed kurta with trouser and dupattaKurt...  Garment Measurements in   \n",
       "4  Black and green printed straight kurta, has a ...  Garment Measurements in   \n",
       "\n",
       "              Neck Occasion Print or Pattern Type          Sleeve Length  \\\n",
       "0  Mandarin Collar  Festive         Ethnic Motifs  Three-Quarter Sleeves   \n",
       "1      Square Neck   Fusion                 Solid             Sleeveless   \n",
       "2       Round Neck    Daily         Ethnic Motifs  Three-Quarter Sleeves   \n",
       "3       Round Neck    Daily         Ethnic Motifs  Three-Quarter Sleeves   \n",
       "4       Round Neck    Daily         Ethnic Motifs  Three-Quarter Sleeves   \n",
       "\n",
       "  Sustainable  ...     Weave Type  Fabric   Hemline       Length  Pattern  \\\n",
       "0     Regular  ...  Machine Weave       0         0            0        0   \n",
       "1     Regular  ...  Machine Weave       0         0            0        0   \n",
       "2     Regular  ...  Machine Weave       0         0            0        0   \n",
       "3     Regular  ...  Machine Weave       0         0            0        0   \n",
       "4           0  ...  Machine Weave  Cotton  Straight  Calf Length  Printed   \n",
       "\n",
       "  Center Front Open Closure Surface Styling Type Sport  \n",
       "0                 0       0               0    0     0  \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",
       "[5 rows x 21 columns]"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "fashion_df = pd.read_csv('/Users/nitanshjain/Documents/Projects/Recommeder_System/data/cleaned_data.csv')\n",
    "fashion_df = fashion_df.drop(columns=['Unnamed: 0', 'p_id', 'price', 'img', 'ratingCount', 'avg_rating', 'p_attributes', 'Body Shape ID', 'Character', 'Number of Pockets', 'Multipack Set', 'Technology', 'Hood', 'Pocket'])\n",
    "print(fashion_df.columns)\n",
    "print(fashion_df.isna().sum())\n",
    "# print(fashion_df.dropna(inplace=True))\n",
    "print(fashion_df.shape)\n",
    "fashion_df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Text PreProcessing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [],
   "source": [
    "def contractions_handling(text):\n",
    "    text = contractions.fix(text)\n",
    "    return text\n",
    "\n",
    "def replacing_abbr(text, dictry):\n",
    "    tokenizer = TweetTokenizer()\n",
    "    tokens = tokenizer.tokenize(text)\n",
    "    x = [dictry[word] if word in dictry.keys() else word for word in tokens]\n",
    "    text = ' '.join(x)\n",
    "    return text\n",
    "\n",
    "def removing_punctuations(text):\n",
    "    # text = text.translate(str.maketrans('', '', string.punctuation))\n",
    "    text = text.replace('-', ' ').replace(':', ' ')\n",
    "    text = re.sub(r'[^\\w\\s]', ' ', text)\n",
    "    text = text.lower()\n",
    "    return text\n",
    "\n",
    "def removing_stopwords(text):\n",
    "    tokenizer = TweetTokenizer()\n",
    "    tokens = tokenizer.tokenize(text)\n",
    "    x = [word for word in tokens if word not in stopwords.words('english')]\n",
    "    text = ' '.join(x)\n",
    "    return text\n",
    "\n",
    "def convert_num_words(text):\n",
    "    re_results = re.findall('(\\d+(st|nd|rd|th))', text)\n",
    "    for entire_result, suffix in re_results:\n",
    "        num = int(entire_result[:-len(suffix)])\n",
    "        entire_result = \" \" + entire_result\n",
    "        num_word = num2words.num2words(num, ordinal=True)\n",
    "        text = text.replace(entire_result, \" \" + num_word)\n",
    "    \n",
    "    tokenizer = TweetTokenizer()\n",
    "    tokens = tokenizer.tokenize(text)\n",
    "    tokens = [num2words.num2words(word) if word.isdigit() else word for word in tokens]\n",
    "    text = ' '.join(tokens)\n",
    "    \n",
    "    return text\n",
    "\n",
    "def removing_char_less_3(text):\n",
    "    tokenizer = TweetTokenizer()\n",
    "    tokens = tokenizer.tokenize(text)\n",
    "    tokens = [word for word in tokens if len(word) >= 3]\n",
    "    text = ' '.join(tokens)\n",
    "    return text\n",
    "\n",
    "def preprocess_text(text):\n",
    "    text = convert_num_words(text)\n",
    "    text = contractions_handling(text)\n",
    "    # text = replacing_abbr(text, abbr_dict)\n",
    "    text = removing_punctuations(text)\n",
    "    text = removing_stopwords(text)\n",
    "    text = removing_char_less_3(text)\n",
    "    return text"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "fashion_df['processed_description'] = fashion_df['description'].apply(lambda x: preprocess_text(x))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Generating Embeddings"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(14309, 2)\n",
      "name                     0\n",
      "processed_description    0\n",
      "dtype: int64\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>name</th>\n",
       "      <th>processed_description</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Khushal K Women Black Ethnic Motifs Printed Ku...</td>\n",
       "      <td>black printed kurta palazzos dupatta kurta des...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>InWeave Women Orange Solid Kurta with Palazzos...</td>\n",
       "      <td>orange solid kurta palazzos dupattakurta desig...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Anubhutee Women Navy Blue Ethnic Motifs Embroi...</td>\n",
       "      <td>navy blue embroidered kurta trousers dupatta k...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Nayo Women Red Floral Printed Kurta With Trous...</td>\n",
       "      <td>red printed kurta trouser dupattakurta design ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>AHIKA Women Black &amp; Green Printed Straight Kurta</td>\n",
       "      <td>black green printed straight kurta nitched rou...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                name  \\\n",
       "0  Khushal K Women Black Ethnic Motifs Printed Ku...   \n",
       "1  InWeave Women Orange Solid Kurta with Palazzos...   \n",
       "2  Anubhutee Women Navy Blue Ethnic Motifs Embroi...   \n",
       "3  Nayo Women Red Floral Printed Kurta With Trous...   \n",
       "4   AHIKA Women Black & Green Printed Straight Kurta   \n",
       "\n",
       "                               processed_description  \n",
       "0  black printed kurta palazzos dupatta kurta des...  \n",
       "1  orange solid kurta palazzos dupattakurta desig...  \n",
       "2  navy blue embroidered kurta trousers dupatta k...  \n",
       "3  red printed kurta trouser dupattakurta design ...  \n",
       "4  black green printed straight kurta nitched rou...  "
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "recommend_df = fashion_df[['name', 'processed_description']].copy()\n",
    "print(recommend_df.shape)\n",
    "print(recommend_df.isna().sum())\n",
    "recommend_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(14309, 10762)"
      ]
     },
     "execution_count": 84,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tfidf_obj = TfidfVectorizer()\n",
    "description_tfidf = tfidf_obj.fit_transform(recommend_df.processed_description)\n",
    "description_tfidf.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'numpy.ndarray'>\n"
     ]
    }
   ],
   "source": [
    "cosine_sim = cosine_similarity(description_tfidf, description_tfidf)\n",
    "print(type(cosine_sim))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [],
   "source": [
    "indices = pd.Series(recommend_df.name)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 117,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(14310, 10762)\n",
      "[7324, 627, 9281, 9344, 9623]\n",
      "7324    0.513636\n",
      "627     0.456661\n",
      "9281    0.454681\n",
      "9344    0.452017\n",
      "9623    0.450200\n",
      "dtype: float64\n",
      "['Soch Women Yellow Georgette Solid Round Neck Suit Set', 'Bhama Couture Women Mustard Yellow Yoke Design Kurta with Palazzos', 'Soch Yellow & Black Embroidered Art Silk Unstitched Dress Material', 'INDIAN HERITAGE Yellow & Red Printed Silk Crepe Unstitched Dress Material', 'Stylee LIFESTYLE Yellow & Pink Pure Silk Unstitched Dress Material']\n"
     ]
    }
   ],
   "source": [
    "import scipy.sparse as sp\n",
    "\n",
    "def recommend(title, recommend=recommend_df, df_tfidf=description_tfidf, tfidf_obj=tfidf_obj):\n",
    "    \n",
    "    title_series = pd.Series([title])\n",
    "    indices = pd.Series(recommend['name'])\n",
    "    indices = indices._append(title_series)\n",
    "    \n",
    "    \n",
    "    title_embeddings = tfidf_obj.transform([title])\n",
    "    title_embeddings = title_embeddings.toarray()\n",
    "    df_tfidf = df_tfidf.toarray()\n",
    "    \n",
    "    new_description_tfidf = np.vstack((df_tfidf, title_embeddings))\n",
    "    print(new_description_tfidf.shape)\n",
    "    cosine_sim = cosine_similarity(new_description_tfidf, new_description_tfidf)\n",
    "    \n",
    "    recommended_movies = []\n",
    "    idx = len(indices)-1 \n",
    "    score_series = pd.Series(cosine_sim[idx]).sort_values(ascending = False)   \n",
    "    top_5_indices = list(score_series.iloc[1:6].index)  \n",
    "    print(top_5_indices)\n",
    "    print(score_series.iloc[1:6])\n",
    "    for i in top_5_indices:   # to append the titles of top 10 similar movies to the recommended_movies list\n",
    "        recommended_movies.append(list(recommend['name'])[i])\n",
    "        \n",
    "        \n",
    "    return recommended_movies\n",
    "    \n",
    "    \n",
    "\n",
    "print(recommend('yellow kurta for diwali'))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Recommeder_System-28XCvFNb",
   "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": "3.10.4"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}