{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "QU13Xg9n0bSM"
   },
   "source": [
    "# Retail Product Recommendations using word2vec\n",
    "> Creating a system that automatically recommends a certain number of products to the consumers on an E-commerce website based on the past purchase behavior of the consumers.\n",
    "\n",
    "- toc: true\n",
    "- badges: true\n",
    "- comments: true\n",
    "- categories: [sequence, retail]\n",
    "- image: "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "VwgbuHwVtshy"
   },
   "source": [
    "A person involved in sports-related activities might have an online buying pattern similar to this:"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "NQLbbKfmtn_O"
   },
   "source": [
    "![image.png](data:image/png;base64,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)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "2tYh4SrWtynJ"
   },
   "source": [
    "If we can represent each of these products by a vector, then we can easily find similar products. So, if a user is checking out a product online, then we can easily recommend him/her similar products by using the vector similarity score between the products."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "executionInfo": {
     "elapsed": 1427,
     "status": "ok",
     "timestamp": 1619252390068,
     "user": {
      "displayName": "sparsh agarwal",
      "photoUrl": "",
      "userId": "00322518567794762549"
     },
     "user_tz": -330
    },
    "id": "W0wI5j74nc9W"
   },
   "outputs": [],
   "source": [
    "#hide\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import random\n",
    "from tqdm import tqdm\n",
    "from gensim.models import Word2Vec \n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "\n",
    "import warnings;\n",
    "warnings.filterwarnings('ignore')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "cDi5Gu8Ou7bb"
   },
   "source": [
    "## Data gathering and understanding"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 3274,
     "status": "ok",
     "timestamp": 1619252391927,
     "user": {
      "displayName": "sparsh agarwal",
      "photoUrl": "",
      "userId": "00322518567794762549"
     },
     "user_tz": -330
    },
    "id": "TlW0pTzGniGo",
    "outputId": "0392f779-4c42-4e9e-b096-51ee6487b53d"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "--2021-04-24 08:19:50--  https://archive.ics.uci.edu/ml/machine-learning-databases/00352/Online%20Retail.xlsx\n",
      "Resolving archive.ics.uci.edu (archive.ics.uci.edu)... 128.195.10.252\n",
      "Connecting to archive.ics.uci.edu (archive.ics.uci.edu)|128.195.10.252|:443... connected.\n",
      "HTTP request sent, awaiting response... 200 OK\n",
      "Length: 23715344 (23M) [application/x-httpd-php]\n",
      "Saving to: ‘Online Retail.xlsx’\n",
      "\n",
      "Online Retail.xlsx  100%[===================>]  22.62M  22.7MB/s    in 1.0s    \n",
      "\n",
      "2021-04-24 08:19:51 (22.7 MB/s) - ‘Online Retail.xlsx’ saved [23715344/23715344]\n",
      "\n"
     ]
    }
   ],
   "source": [
    "#hide-output\n",
    "!wget https://archive.ics.uci.edu/ml/machine-learning-databases/00352/Online%20Retail.xlsx"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 204
    },
    "executionInfo": {
     "elapsed": 45709,
     "status": "ok",
     "timestamp": 1619252434373,
     "user": {
      "displayName": "sparsh agarwal",
      "photoUrl": "",
      "userId": "00322518567794762549"
     },
     "user_tz": -330
    },
    "id": "z_Kt8wtZnjRm",
    "outputId": "e30cdfd6-fd16-48ab-fa22-283fdb3d2578"
   },
   "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>InvoiceNo</th>\n",
       "      <th>StockCode</th>\n",
       "      <th>Description</th>\n",
       "      <th>Quantity</th>\n",
       "      <th>InvoiceDate</th>\n",
       "      <th>UnitPrice</th>\n",
       "      <th>CustomerID</th>\n",
       "      <th>Country</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>536365</td>\n",
       "      <td>85123A</td>\n",
       "      <td>WHITE HANGING HEART T-LIGHT HOLDER</td>\n",
       "      <td>6</td>\n",
       "      <td>2010-12-01 08:26:00</td>\n",
       "      <td>2.55</td>\n",
       "      <td>17850.0</td>\n",
       "      <td>United Kingdom</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>536365</td>\n",
       "      <td>71053</td>\n",
       "      <td>WHITE METAL LANTERN</td>\n",
       "      <td>6</td>\n",
       "      <td>2010-12-01 08:26:00</td>\n",
       "      <td>3.39</td>\n",
       "      <td>17850.0</td>\n",
       "      <td>United Kingdom</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>536365</td>\n",
       "      <td>84406B</td>\n",
       "      <td>CREAM CUPID HEARTS COAT HANGER</td>\n",
       "      <td>8</td>\n",
       "      <td>2010-12-01 08:26:00</td>\n",
       "      <td>2.75</td>\n",
       "      <td>17850.0</td>\n",
       "      <td>United Kingdom</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>536365</td>\n",
       "      <td>84029G</td>\n",
       "      <td>KNITTED UNION FLAG HOT WATER BOTTLE</td>\n",
       "      <td>6</td>\n",
       "      <td>2010-12-01 08:26:00</td>\n",
       "      <td>3.39</td>\n",
       "      <td>17850.0</td>\n",
       "      <td>United Kingdom</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>536365</td>\n",
       "      <td>84029E</td>\n",
       "      <td>RED WOOLLY HOTTIE WHITE HEART.</td>\n",
       "      <td>6</td>\n",
       "      <td>2010-12-01 08:26:00</td>\n",
       "      <td>3.39</td>\n",
       "      <td>17850.0</td>\n",
       "      <td>United Kingdom</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  InvoiceNo StockCode  ... CustomerID         Country\n",
       "0    536365    85123A  ...    17850.0  United Kingdom\n",
       "1    536365     71053  ...    17850.0  United Kingdom\n",
       "2    536365    84406B  ...    17850.0  United Kingdom\n",
       "3    536365    84029G  ...    17850.0  United Kingdom\n",
       "4    536365    84029E  ...    17850.0  United Kingdom\n",
       "\n",
       "[5 rows x 8 columns]"
      ]
     },
     "execution_count": 3,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.read_excel('Online Retail.xlsx')\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "KJHexOy0oPHl"
   },
   "source": [
    "Given below is the description of the fields in this dataset:\n",
    "\n",
    "1. __InvoiceNo:__ Invoice number, a unique number assigned to each transaction.\n",
    "\n",
    "2. __StockCode:__ Product/item code. a unique number assigned to each distinct product.\n",
    "\n",
    "3. __Description:__ Product description\n",
    "\n",
    "4. __Quantity:__ The quantities of each product per transaction.\n",
    "\n",
    "5. __InvoiceDate:__ Invoice Date and time. The day and time when each transaction was generated.\n",
    "\n",
    "6. __CustomerID:__ Customer number, a unique number assigned to each customer."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "h8BwXsF5ox--"
   },
   "source": [
    "## Data Preprocessing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 45703,
     "status": "ok",
     "timestamp": 1619252434375,
     "user": {
      "displayName": "sparsh agarwal",
      "photoUrl": "",
      "userId": "00322518567794762549"
     },
     "user_tz": -330
    },
    "id": "F58y_bU6nypT",
    "outputId": "26f4b722-e131-46af-9dde-3cf388986882"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "InvoiceNo           0\n",
       "StockCode           0\n",
       "Description      1454\n",
       "Quantity            0\n",
       "InvoiceDate         0\n",
       "UnitPrice           0\n",
       "CustomerID     135080\n",
       "Country             0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 4,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# check for missing values\n",
    "df.isnull().sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "JM04nhUAot7Y"
   },
   "source": [
    "Since we have sufficient data, we will drop all the rows with missing values."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 45696,
     "status": "ok",
     "timestamp": 1619252434376,
     "user": {
      "displayName": "sparsh agarwal",
      "photoUrl": "",
      "userId": "00322518567794762549"
     },
     "user_tz": -330
    },
    "id": "uqEjaaTKorZ4",
    "outputId": "061df95f-dfbf-4fda-fd6c-f3f329df87ed"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "InvoiceNo      0\n",
       "StockCode      0\n",
       "Description    0\n",
       "Quantity       0\n",
       "InvoiceDate    0\n",
       "UnitPrice      0\n",
       "CustomerID     0\n",
       "Country        0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 5,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# remove missing values\n",
    "df.dropna(inplace=True)\n",
    "\n",
    "# again check missing values\n",
    "df.isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "executionInfo": {
     "elapsed": 46353,
     "status": "ok",
     "timestamp": 1619252435036,
     "user": {
      "displayName": "sparsh agarwal",
      "photoUrl": "",
      "userId": "00322518567794762549"
     },
     "user_tz": -330
    },
    "id": "RUib7vkCpd1E"
   },
   "outputs": [],
   "source": [
    "# Convert the StockCode to string datatype\n",
    "df['StockCode']= df['StockCode'].astype(str)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 46347,
     "status": "ok",
     "timestamp": 1619252435036,
     "user": {
      "displayName": "sparsh agarwal",
      "photoUrl": "",
      "userId": "00322518567794762549"
     },
     "user_tz": -330
    },
    "id": "wPI7YWkQo6wu",
    "outputId": "e751a6d9-6ba2-42a7-8a4f-1aaee616f2b4"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "4372"
      ]
     },
     "execution_count": 7,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Check out the number of unique customers in our dataset\n",
    "customers = df[\"CustomerID\"].unique().tolist()\n",
    "len(customers)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "ED5y4TDxpOfL"
   },
   "source": [
    "There are 4,372 customers in our dataset. For each of these customers we will extract their buying history. In other words, we can have 4,372 sequences of purchases."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "Aogi1piGvEGl"
   },
   "source": [
    "## Data Preparation"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "F_WFU-rTvJ_l"
   },
   "source": [
    "It is a good practice to set aside a small part of the dataset for validation purpose. Therefore, we will use data of 90% of the customers to create word2vec embeddings. Let's split the data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "executionInfo": {
     "elapsed": 46345,
     "status": "ok",
     "timestamp": 1619252435037,
     "user": {
      "displayName": "sparsh agarwal",
      "photoUrl": "",
      "userId": "00322518567794762549"
     },
     "user_tz": -330
    },
    "id": "WxWaZx3zpPFW"
   },
   "outputs": [],
   "source": [
    "# shuffle customer ID's\n",
    "random.shuffle(customers)\n",
    "\n",
    "# extract 90% of customer ID's\n",
    "customers_train = [customers[i] for i in range(round(0.9*len(customers)))]\n",
    "\n",
    "# split data into train and validation set\n",
    "train_df = df[df['CustomerID'].isin(customers_train)]\n",
    "validation_df = df[~df['CustomerID'].isin(customers_train)]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "ul94yT2cqJ38"
   },
   "source": [
    "Let's create sequences of purchases made by the customers in the dataset for both the train and validation set."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 51786,
     "status": "ok",
     "timestamp": 1619252440487,
     "user": {
      "displayName": "sparsh agarwal",
      "photoUrl": "",
      "userId": "00322518567794762549"
     },
     "user_tz": -330
    },
    "id": "uhIwWEe-qGwK",
    "outputId": "6de33b11-1871-4fea-e60e-56e4cb03db7a"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 3935/3935 [00:05<00:00, 664.97it/s]\n"
     ]
    }
   ],
   "source": [
    "# list to capture purchase history of the customers\n",
    "purchases_train = []\n",
    "\n",
    "# populate the list with the product codes\n",
    "for i in tqdm(customers_train):\n",
    "    temp = train_df[train_df[\"CustomerID\"] == i][\"StockCode\"].tolist()\n",
    "    purchases_train.append(temp)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 52224,
     "status": "ok",
     "timestamp": 1619252440935,
     "user": {
      "displayName": "sparsh agarwal",
      "photoUrl": "",
      "userId": "00322518567794762549"
     },
     "user_tz": -330
    },
    "id": "VGT9oyVeqhky",
    "outputId": "8198f6f0-9649-4214-ef55-20576f4a6e9f"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 437/437 [00:00<00:00, 1006.50it/s]\n"
     ]
    }
   ],
   "source": [
    "# list to capture purchase history of the customers\n",
    "purchases_val = []\n",
    "\n",
    "# populate the list with the product codes\n",
    "for i in tqdm(validation_df['CustomerID'].unique()):\n",
    "    temp = validation_df[validation_df[\"CustomerID\"] == i][\"StockCode\"].tolist()\n",
    "    purchases_val.append(temp)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "AgDLwI_4q4Fm"
   },
   "source": [
    "## Build word2vec Embeddings for Products"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 106693,
     "status": "ok",
     "timestamp": 1619252495414,
     "user": {
      "displayName": "sparsh agarwal",
      "photoUrl": "",
      "userId": "00322518567794762549"
     },
     "user_tz": -330
    },
    "id": "rr_tHmmuqu24",
    "outputId": "dcedddc7-d410-4d0c-bd3a-79f6022f42ef"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(3657318, 3696290)"
      ]
     },
     "execution_count": 11,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# train word2vec model\n",
    "model = Word2Vec(window = 10, sg = 1, hs = 0,\n",
    "                 negative = 10, # for negative sampling\n",
    "                 alpha=0.03, min_alpha=0.0007,\n",
    "                 seed = 14)\n",
    "\n",
    "model.build_vocab(purchases_train, progress_per=200)\n",
    "\n",
    "model.train(purchases_train, total_examples = model.corpus_count, \n",
    "            epochs=10, report_delay=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "executionInfo": {
     "elapsed": 106690,
     "status": "ok",
     "timestamp": 1619252495414,
     "user": {
      "displayName": "sparsh agarwal",
      "photoUrl": "",
      "userId": "00322518567794762549"
     },
     "user_tz": -330
    },
    "id": "_CwTzmNqq_lQ"
   },
   "outputs": [],
   "source": [
    "# save word2vec model\n",
    "model.save(\"word2vec_2.model\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "HBbudJ7hrDw0"
   },
   "source": [
    "As we do not plan to train the model any further, we are calling init_sims(), which will make the model much more memory-efficient"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "executionInfo": {
     "elapsed": 106689,
     "status": "ok",
     "timestamp": 1619252495415,
     "user": {
      "displayName": "sparsh agarwal",
      "photoUrl": "",
      "userId": "00322518567794762549"
     },
     "user_tz": -330
    },
    "id": "UGYK8p1xrAJy"
   },
   "outputs": [],
   "source": [
    "model.init_sims(replace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 106680,
     "status": "ok",
     "timestamp": 1619252495416,
     "user": {
      "displayName": "sparsh agarwal",
      "photoUrl": "",
      "userId": "00322518567794762549"
     },
     "user_tz": -330
    },
    "id": "M_AbUfMOrGsI",
    "outputId": "66d6d641-2dfb-4129-b27c-6eeafffd07a2"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Word2Vec(vocab=3153, size=100, alpha=0.03)\n"
     ]
    }
   ],
   "source": [
    "print(model)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "HKFT2SECrMY1"
   },
   "source": [
    "Now we will extract the vectors of all the words in our vocabulary and store it in one place for easy access"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 106664,
     "status": "ok",
     "timestamp": 1619252495416,
     "user": {
      "displayName": "sparsh agarwal",
      "photoUrl": "",
      "userId": "00322518567794762549"
     },
     "user_tz": -330
    },
    "id": "odn6f3rorG7T",
    "outputId": "f8c06bf9-1ba4-48f9-e491-1fb2369fdadc"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(3153, 100)"
      ]
     },
     "execution_count": 15,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# extract all vectors\n",
    "X = model[model.wv.vocab]\n",
    "\n",
    "X.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "VQr5YRjIrTC2"
   },
   "source": [
    "## Visualize word2vec Embeddings"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "EwW0xbwkrVoB"
   },
   "source": [
    "It is always quite helpful to visualize the embeddings that you have created. Over here we have 100 dimensional embeddings. We can't even visualize 4 dimensions let alone 100. Therefore, we are going to reduce the dimensions of the product embeddings from 100 to 2 by using the UMAP algorithm, it is used for dimensionality reduction."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 109415,
     "status": "ok",
     "timestamp": 1619252498177,
     "user": {
      "displayName": "sparsh agarwal",
      "photoUrl": "",
      "userId": "00322518567794762549"
     },
     "user_tz": -330
    },
    "id": "A7XmYq1Eragv",
    "outputId": "9dfed0d8-4a6c-4cd5-bf14-9bbe0e8248f2"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Requirement already satisfied: umap-learn in /usr/local/lib/python3.7/dist-packages (0.5.1)\n",
      "Requirement already satisfied: numba>=0.49 in /usr/local/lib/python3.7/dist-packages (from umap-learn) (0.51.2)\n",
      "Requirement already satisfied: numpy>=1.17 in /usr/local/lib/python3.7/dist-packages (from umap-learn) (1.19.5)\n",
      "Requirement already satisfied: pynndescent>=0.5 in /usr/local/lib/python3.7/dist-packages (from umap-learn) (0.5.2)\n",
      "Requirement already satisfied: scipy>=1.0 in /usr/local/lib/python3.7/dist-packages (from umap-learn) (1.4.1)\n",
      "Requirement already satisfied: scikit-learn>=0.22 in /usr/local/lib/python3.7/dist-packages (from umap-learn) (0.22.2.post1)\n",
      "Requirement already satisfied: setuptools in /usr/local/lib/python3.7/dist-packages (from numba>=0.49->umap-learn) (56.0.0)\n",
      "Requirement already satisfied: llvmlite<0.35,>=0.34.0.dev0 in /usr/local/lib/python3.7/dist-packages (from numba>=0.49->umap-learn) (0.34.0)\n",
      "Requirement already satisfied: joblib>=0.11 in /usr/local/lib/python3.7/dist-packages (from pynndescent>=0.5->umap-learn) (1.0.1)\n"
     ]
    }
   ],
   "source": [
    "#hide\n",
    "!pip install umap-learn"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 54
    },
    "executionInfo": {
     "elapsed": 150522,
     "status": "ok",
     "timestamp": 1619252539293,
     "user": {
      "displayName": "sparsh agarwal",
      "photoUrl": "",
      "userId": "00322518567794762549"
     },
     "user_tz": -330
    },
    "id": "Y8wKkbRQrQAy",
    "outputId": "0b928341-1790-411d-fb91-a824f3c05264"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x648 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light",
      "tags": []
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#collapse\n",
    "import umap\n",
    "\n",
    "cluster_embedding = umap.UMAP(n_neighbors=30, min_dist=0.0,\n",
    "                              n_components=2, random_state=42).fit_transform(X)\n",
    "\n",
    "plt.figure(figsize=(10,9))\n",
    "plt.scatter(cluster_embedding[:, 0], cluster_embedding[:, 1], s=3, cmap='Spectral');"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "wNQ9hotsr1eC"
   },
   "source": [
    "Every dot in this plot is a product. As you can see, there are several tiny clusters of these datapoints. These are groups of similar products."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "5_KBjh2Jr-Qg"
   },
   "source": [
    "## Generate and validate recommendations"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "he6azpQysC_M"
   },
   "source": [
    "We are finally ready with the word2vec embeddings for every product in our online retail dataset. Now our next step is to suggest similar products for a certain product or a product's vector. \n",
    "\n",
    "Let's first create a product-ID and product-description dictionary to easily map a product's description to its ID and vice versa."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "executionInfo": {
     "elapsed": 150520,
     "status": "ok",
     "timestamp": 1619252539294,
     "user": {
      "displayName": "sparsh agarwal",
      "photoUrl": "",
      "userId": "00322518567794762549"
     },
     "user_tz": -330
    },
    "id": "sckVfOMHrYHp"
   },
   "outputs": [],
   "source": [
    "products = train_df[[\"StockCode\", \"Description\"]]\n",
    "\n",
    "# remove duplicates\n",
    "products.drop_duplicates(inplace=True, subset='StockCode', keep=\"last\")\n",
    "\n",
    "# create product-ID and product-description dictionary\n",
    "products_dict = products.groupby('StockCode')['Description'].apply(list).to_dict()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 150514,
     "status": "ok",
     "timestamp": 1619252539296,
     "user": {
      "displayName": "sparsh agarwal",
      "photoUrl": "",
      "userId": "00322518567794762549"
     },
     "user_tz": -330
    },
    "id": "Ldj0ew7UsOiw",
    "outputId": "e03b87e5-8f3c-466e-a85c-ec004746efdc"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['RED WOOLLY HOTTIE WHITE HEART.']"
      ]
     },
     "execution_count": 19,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# test the dictionary\n",
    "products_dict['84029E']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "eneoWxrwsSjt"
   },
   "source": [
    "We have defined the function below. It will take a product's vector (n) as input and return top 6 similar products."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "executionInfo": {
     "elapsed": 150514,
     "status": "ok",
     "timestamp": 1619252539298,
     "user": {
      "displayName": "sparsh agarwal",
      "photoUrl": "",
      "userId": "00322518567794762549"
     },
     "user_tz": -330
    },
    "id": "TmgZr0c2sPuz"
   },
   "outputs": [],
   "source": [
    "#hide\n",
    "def similar_products(v, n = 6):\n",
    "    \n",
    "    # extract most similar products for the input vector\n",
    "    ms = model.similar_by_vector(v, topn= n+1)[1:]\n",
    "    \n",
    "    # extract name and similarity score of the similar products\n",
    "    new_ms = []\n",
    "    for j in ms:\n",
    "        pair = (products_dict[j[0]][0], j[1])\n",
    "        new_ms.append(pair)\n",
    "        \n",
    "    return new_ms        "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "H3ygmMQHsZRA"
   },
   "source": [
    "Let's try out our function by passing the vector of the product '90019A' ('SILVER M.O.P ORBIT BRACELET')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 150506,
     "status": "ok",
     "timestamp": 1619252539299,
     "user": {
      "displayName": "sparsh agarwal",
      "photoUrl": "",
      "userId": "00322518567794762549"
     },
     "user_tz": -330
    },
    "id": "pWnP50EVsWEj",
    "outputId": "bec2fe21-2259-4ae6-a30d-5cb4667d53d7"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[('SILVER M.O.P ORBIT DROP EARRINGS', 0.7879312634468079),\n",
       " ('AMBER DROP EARRINGS W LONG BEADS', 0.7682332992553711),\n",
       " ('JADE DROP EARRINGS W FILIGREE', 0.761816143989563),\n",
       " ('DROP DIAMANTE EARRINGS PURPLE', 0.7489826679229736),\n",
       " ('SILVER LARIAT BLACK STONE EARRINGS', 0.7389366626739502),\n",
       " ('WHITE VINT ART DECO CRYSTAL NECKLAC', 0.7352254390716553)]"
      ]
     },
     "execution_count": 21,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "similar_products(model['90019A'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "aH-2nt_ishkM"
   },
   "source": [
    "Cool! The results are pretty relevant and match well with the input product. However, this output is based on the vector of a single product only. What if we want recommend a user products based on the multiple purchases he or she has made in the past?\n",
    "\n",
    "One simple solution is to take average of all the vectors of the products he has bought so far and use this resultant vector to find similar products. For that we will use the function below that takes in a list of product ID's and gives out a 100 dimensional vector which is mean of vectors of the products in the input list."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "executionInfo": {
     "elapsed": 150504,
     "status": "ok",
     "timestamp": 1619252539299,
     "user": {
      "displayName": "sparsh agarwal",
      "photoUrl": "",
      "userId": "00322518567794762549"
     },
     "user_tz": -330
    },
    "id": "MqtpgpqFsauG"
   },
   "outputs": [],
   "source": [
    "#collapse\n",
    "def aggregate_vectors(products):\n",
    "    product_vec = []\n",
    "    for i in products:\n",
    "        try:\n",
    "            product_vec.append(model[i])\n",
    "        except KeyError:\n",
    "            continue\n",
    "        \n",
    "    return np.mean(product_vec, axis=0)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "mXPqS4h0sojc"
   },
   "source": [
    "If you can recall, we have already created a separate list of purchase sequences for validation purpose. Now let's make use of that."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 151381,
     "status": "ok",
     "timestamp": 1619252540183,
     "user": {
      "displayName": "sparsh agarwal",
      "photoUrl": "",
      "userId": "00322518567794762549"
     },
     "user_tz": -330
    },
    "id": "KfI_RrLZsn8W",
    "outputId": "929f549e-1621-474d-9c96-f4012a1aa5f9"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "28"
      ]
     },
     "execution_count": 23,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#hide\n",
    "len(purchases_val[0])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "nuaqWGa4ssZr"
   },
   "source": [
    "The length of the first list of products purchased by a user is 314. We will pass this products' sequence of the validation set to the function aggregate_vectors."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 151376,
     "status": "ok",
     "timestamp": 1619252540184,
     "user": {
      "displayName": "sparsh agarwal",
      "photoUrl": "",
      "userId": "00322518567794762549"
     },
     "user_tz": -330
    },
    "id": "OdurKB3ysp6c",
    "outputId": "c2635b4c-584c-47a0-e436-f4175fc87dad"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(100,)"
      ]
     },
     "execution_count": 24,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#hide\n",
    "aggregate_vectors(purchases_val[0]).shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "gwaXOQSus9ya"
   },
   "source": [
    "Well, the function has returned an array of 100 dimension. It means the function is working fine. Now we can use this result to get the most similar products. Let's do it."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 151371,
     "status": "ok",
     "timestamp": 1619252540186,
     "user": {
      "displayName": "sparsh agarwal",
      "photoUrl": "",
      "userId": "00322518567794762549"
     },
     "user_tz": -330
    },
    "id": "PKTVsoc3s7ZK",
    "outputId": "ec1f2e74-74bc-4035-e8f6-b01b95b35f4d"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[('WHITE SPOT BLUE CERAMIC DRAWER KNOB', 0.6860978603363037),\n",
       " ('RED SPOT CERAMIC DRAWER KNOB', 0.6785424947738647),\n",
       " ('BLUE STRIPE CERAMIC DRAWER KNOB', 0.6783121824264526),\n",
       " ('BLUE SPOT CERAMIC DRAWER KNOB', 0.6738985776901245),\n",
       " ('CLEAR DRAWER KNOB ACRYLIC EDWARDIAN', 0.6731897592544556),\n",
       " ('RED STRIPE CERAMIC DRAWER KNOB', 0.6667704582214355)]"
      ]
     },
     "execution_count": 25,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "similar_products(aggregate_vectors(purchases_val[0]))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "MRsPrkMotDaZ"
   },
   "source": [
    "As it turns out, our system has recommended 6 products based on the entire purchase history of a user. Moreover, if you want to get products suggestions based on the last few purchases only then also you can use the same set of functions.\n",
    "\n",
    "Below we are giving only the last 10 products purchased as input."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 151364,
     "status": "ok",
     "timestamp": 1619252540187,
     "user": {
      "displayName": "sparsh agarwal",
      "photoUrl": "",
      "userId": "00322518567794762549"
     },
     "user_tz": -330
    },
    "id": "OUqyW1-Ns_u2",
    "outputId": "c903bdf6-ac73-4a53-eaa6-8987a67cc8fb"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[('BLUE SPOT CERAMIC DRAWER KNOB', 0.7394766807556152),\n",
       " ('RED SPOT CERAMIC DRAWER KNOB', 0.7364704012870789),\n",
       " ('WHITE SPOT BLUE CERAMIC DRAWER KNOB', 0.7347637414932251),\n",
       " ('ASSORTED COLOUR BIRD ORNAMENT', 0.7345550060272217),\n",
       " ('RED STRIPE CERAMIC DRAWER KNOB', 0.7305896878242493),\n",
       " ('WHITE SPOT RED CERAMIC DRAWER KNOB', 0.6979628801345825)]"
      ]
     },
     "execution_count": 26,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "similar_products(aggregate_vectors(purchases_val[0][-10:]))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "59iEesnTzNHq"
   },
   "source": [
    "## References\n",
    "\n",
    "- [https://www.analyticsvidhya.com/blog/2019/07/how-to-build-recommendation-system-word2vec-python/](https://www.analyticsvidhya.com/blog/2019/07/how-to-build-recommendation-system-word2vec-python/)\n",
    "- [https://mccormickml.com/2018/06/15/applying-word2vec-to-recommenders-and-advertising/](https://mccormickml.com/2018/06/15/applying-word2vec-to-recommenders-and-advertising/)\n",
    "- [https://www.analyticsinsight.net/building-recommendation-system-using-item2vec/](https://www.analyticsinsight.net/building-recommendation-system-using-item2vec/)\n",
    "- [https://towardsdatascience.com/using-word2vec-for-music-recommendations-bb9649ac2484](https://towardsdatascience.com/using-word2vec-for-music-recommendations-bb9649ac2484)\n",
    "- [https://capablemachine.com/2020/06/23/word-embedding/](https://capablemachine.com/2020/06/23/word-embedding/)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "executionInfo": {
     "elapsed": 151363,
     "status": "ok",
     "timestamp": 1619252540188,
     "user": {
      "displayName": "sparsh agarwal",
      "photoUrl": "",
      "userId": "00322518567794762549"
     },
     "user_tz": -330
    },
    "id": "Q--xnW3dtIZy"
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "colab": {
   "collapsed_sections": [],
   "name": "2021-04-24-rec-medium-word2vec.ipynb",
   "provenance": [
    {
     "file_id": "https://gist.github.com/sparsh-ai/ca06820d9124f03e90b33dbe98348639#file-rec-medium-word2vec-ipynb",
     "timestamp": 1619249955624
    }
   ]
  },
  "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": "3.7.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}