Capstone_Project_Erdem_Korhan_Akçay.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    },
    "accelerator": "TPU",
    "gpuClass": "standard"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "source": [
        "# **INTRODUCTION**\n",
        "\n",
        "\n",
        "Recommendation engines have become an increasingly important tool in the world of online commerce. By analyzing a customer's purchase history, these systems are able to make personalized recommendations for products or services that the customer might be interested in. This not only helps to improve the customer experience, but it can also drive sales and increase customer loyalty. In this article, we will explore the basics of recommendation engines and how they can be used to improve the shopping experience for customers. We will also discuss some of the challenges and considerations involved in building and implementing a recommendation engine based on purchase history."
      ],
      "metadata": {
        "id": "oxKYHG3IAP0n"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Trendyol Recommendation Engine\n",
        "\n",
        "Trendyol verisini kullanarak müşterilerin aldıkları ürünleri baz alarak müşterinin yeni alabilecek ürünleri önerim sistemi olacaktır. \n",
        "\n"
      ],
      "metadata": {
        "id": "FUwByo_Yfs63"
      }
    },
    {
      "cell_type": "code",
      "execution_count": 1,
      "metadata": {
        "id": "C4lWhJd7fJuA"
      },
      "outputs": [],
      "source": [
        "import matplotlib\n",
        "import numpy as np\n",
        "import pandas as pd\n",
        "import requests as rq\n",
        "import matplotlib as plt\n",
        "import time\n",
        "import random\n",
        "from docutils.nodes import inline\n",
        "from sklearn.model_selection import train_test_split\n",
        "from wordcloud import WordCloud, STOPWORDS\n",
        "#import turicreate as tc\n",
        "import sys\n",
        "import sqlalchemy\n",
        "\n",
        "sys.path.append(\"..\")\n",
        "\n",
        "# For content based recommendation\n",
        "from sklearn.feature_extraction.text import CountVectorizer\n",
        "from sklearn.metrics.pairwise import cosine_similarity\n",
        "import nltk\n",
        "import re\n",
        "\n",
        "# import warnings\n",
        "\n",
        "# warnings.filterwarnings('ignore')\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "GQQBfNV7z_eb"
      },
      "source": [
        "# **Dataset:**\n",
        "\n",
        "**Column's Names:**\n",
        "\n",
        " 0   partition_date: Satın alma tarihi\n",
        "\n",
        " 1   orderparentid     : Satın alım kodu\n",
        "\n",
        " 2   user_id           : Satın alan kullanıcı\n",
        "\n",
        " 3   productcontentid  : Satın alınan ürün kodu\n",
        "\n",
        " 4   brand_id          : Ürünün marka kodu\n",
        "\n",
        " 5   category_id       : Ürünün dahil olduğu kategori kodu\n",
        "\n",
        " 6   category_name     : Ürünün dahil olduğu kategorinin adı\n",
        "\n",
        " 7   gender            : Ürünün cinsiyeti\n",
        "\n",
        " 8   price             : Ürünün fiyatı\n",
        "\n",
        " 9   color_id          : Ürünün renk kodu\n",
        "\n",
        " 10  business_unit     : Ürünün dahil olduğu genel grubun adı\n",
        "\n",
        " 11  ImageLink         : Ürün görselinin linki\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "### **1. Uploading Dataset**"
      ],
      "metadata": {
        "id": "h9qHIAe6h2a-"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "data = pd.read_csv('https://storage.googleapis.com/ty2020/reco.csv.gz')\n",
        "print(data.head())"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "m520ChsefLHx",
        "outputId": "564cf2ff-8ff7-4b28-b1cf-3dca8454b772"
      },
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "     partition_date  orderparentid  user_id  productcontentid  brand_id  \\\n",
            "0  20.08.2020 06:00      335057357    86386          39328996    919155   \n",
            "1  24.08.2020 10:00      337401625    59469          31903343       121   \n",
            "2  26.08.2020 19:00      338457012    51248          34726400        40   \n",
            "3  22.08.2020 11:00      336681542    29380          32920640      7651   \n",
            "4  20.08.2020 19:00      335736916    68368          39035716      3395   \n",
            "\n",
            "   category_id category_name  gender    price  color_id  \\\n",
            "0          418      Sandalet   Kadın   319.98      14.0   \n",
            "1         1827  Banyo Dolabı  Unisex  1195.56       3.0   \n",
            "2          604       T-Shirt   Kadın    37.99      14.0   \n",
            "3          604       T-Shirt   Kadın    24.99       6.0   \n",
            "4          599         Kazak   Kadın    79.90      16.0   \n",
            "\n",
            "                    business_unit  \\\n",
            "0                 Branded Shoes B   \n",
            "1  Bahçe & Yapı Market & Hırdavat   \n",
            "2                        PL Woman   \n",
            "3                         Kadın A   \n",
            "4                         Kadın A   \n",
            "\n",
            "                                           ImageLink  \n",
            "0  https://cdn.dsmcdn.com//ty1/product/media/imag...  \n",
            "1  https://cdn.dsmcdn.com//assets/product/media/i...  \n",
            "2  https://cdn.dsmcdn.com//assets/product/media/i...  \n",
            "3  https://cdn.dsmcdn.com//assets/product/media/i...  \n",
            "4  https://cdn.dsmcdn.com//ty1/product/media/imag...  \n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "print(data.shape)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Hd-zCnJwhrcS",
        "outputId": "df8515f3-e177-47f6-b74e-f412ca53cdc8"
      },
      "execution_count": 3,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(508228, 12)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "print(data.info())"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "LvFNvPFPhxmn",
        "outputId": "aa4745eb-0e68-4650-dfdb-72822b14766b"
      },
      "execution_count": 4,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "<class 'pandas.core.frame.DataFrame'>\n",
            "RangeIndex: 508228 entries, 0 to 508227\n",
            "Data columns (total 12 columns):\n",
            " #   Column            Non-Null Count   Dtype  \n",
            "---  ------            --------------   -----  \n",
            " 0   partition_date    508228 non-null  object \n",
            " 1   orderparentid     508228 non-null  int64  \n",
            " 2   user_id           508228 non-null  int64  \n",
            " 3   productcontentid  508228 non-null  int64  \n",
            " 4   brand_id          508228 non-null  int64  \n",
            " 5   category_id       508228 non-null  int64  \n",
            " 6   category_name     508228 non-null  object \n",
            " 7   gender            475493 non-null  object \n",
            " 8   price             508228 non-null  float64\n",
            " 9   color_id          375670 non-null  float64\n",
            " 10  business_unit     508228 non-null  object \n",
            " 11  ImageLink         508228 non-null  object \n",
            "dtypes: float64(2), int64(5), object(5)\n",
            "memory usage: 46.5+ MB\n",
            "None\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Let's see how many unique values are in the data set"
      ],
      "metadata": {
        "id": "bcv5lBASCaZP"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# number of unique users\n",
        "num_of_users = len(data[\"user_id\"].unique())\n",
        "print(\"Number of Users:\", num_of_users)\n",
        "\n",
        "# number of unique productcontentid\n",
        "num_of_products = len(data[\"productcontentid\"].unique())\n",
        "print(\"Number of prodcutcontentid\", num_of_products)\n",
        "\n",
        "# number of unique brand_id\n",
        "num_of_brands = len(data[\"brand_id\"].unique())\n",
        "print(\"Number of brands:\", num_of_brands)\n",
        "\n",
        "# number of unique category_id\n",
        "num_of_categories = len(data[\"category_id\"].unique())\n",
        "print(\"Number of categories:\", num_of_categories)\n",
        "\n",
        "# number of unique category_name\n",
        "num_of_category_names = len(data[\"category_name\"].unique())\n",
        "print(\"Number of category_names:\", num_of_category_names)\n",
        "\n",
        "# number of unique orderparentid\n",
        "num_of_orderparentid = len(data[\"orderparentid\"].unique())\n",
        "print(\"Number of Orderparentid:\", num_of_orderparentid)\n",
        "\n",
        "# number of unique gender\n",
        "num_of_genders = data[\"gender\"].unique()\n",
        "print(\"Number of Gender:\", num_of_genders)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "udthJdFDfLE_",
        "outputId": "e5f2ca9c-5a13-4c37-e67c-7bcd195a7956"
      },
      "execution_count": 5,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Number of Users: 108944\n",
            "Number of prodcutcontentid 184092\n",
            "Number of brands: 16838\n",
            "Number of categories: 1879\n",
            "Number of category_names: 1879\n",
            "Number of Orderparentid: 249864\n",
            "Number of Gender: ['Kadın' 'Unisex' nan 'Erkek']\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# **Wordcloud for category_name:**\n",
        "Let's examine the popularity of purchased categories with WordCloud."
      ],
      "metadata": {
        "id": "YbS34i4zhglk"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import matplotlib.pyplot as plt\n",
        "from wordcloud import WordCloud\n",
        "\n",
        "stopwords = set(STOPWORDS)\n",
        "\n",
        "#Creating word_cloud with text as argument in .generate() method\n",
        "word_cloud = WordCloud(collocations = False, background_color = 'white').generate(data['category_name'].to_string())\n",
        "word_cloud = WordCloud(stopwords=stopwords, background_color=\"white\", max_words=1000).generate(data['category_name'].to_string())\n",
        "\n",
        "#Display the generated Word Cloud\n",
        "plt.imshow(word_cloud, interpolation='bilinear')\n",
        "plt.axis(\"off\")\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 198
        },
        "id": "sHEDHES8fLCU",
        "outputId": "23ab3b77-2921-4468-bb29-809858fcb9ea"
      },
      "execution_count": 6,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Data prepration for recommendation algorithms, First, we organize our data by setting the purchase per person at least 4 to achieve a better result.\n"
      ],
      "metadata": {
        "id": "Cwq3I5JmCorr"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "merged = pd.DataFrame(data.groupby(\"user_id\")[\"orderparentid\"].count()).reset_index()\n",
        "cleared_data = data.drop(columns=[\"partition_date\", \"price\"])\n",
        "data_cleaned = merged.merge(cleared_data, on=\"user_id\")\n",
        "final_data = data_cleaned[data_cleaned[\"orderparentid_x\"] > 4]\n",
        "\n",
        "print(final_data.head().to_string())\n",
        "print(final_data.shape)\n",
        "print(final_data.info())"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "iugZkaEpfK_q",
        "outputId": "0a3746e6-8c8b-4a21-ed05-7655561a8405"
      },
      "execution_count": 7,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "    user_id  orderparentid_x  orderparentid_y  productcontentid  brand_id  category_id category_name gender  color_id business_unit                                                                                          ImageLink\n",
            "7       104                5        338745787          44752126      7651          604       T-Shirt  Erkek      17.0       Kadın A      https://cdn.dsmcdn.com//ty7/product/media/images/20200806/8/7251712/79274604/3/3_org_zoom.jpg\n",
            "8       104                5        338745787          36511807      7651          604       T-Shirt  Erkek      17.0       Kadın A  https://cdn.dsmcdn.com//assets/product/media/images/20200306/11/4082924/64991410/1/1_org_zoom.jpg\n",
            "9       104                5        338745787          44751760      7651          604       T-Shirt  Erkek       2.0       Kadın A     https://cdn.dsmcdn.com//ty9/product/media/images/20200901/11/9239398/82986407/1/1_org_zoom.jpg\n",
            "10      104                5        338745787          44750797      7651          604       T-Shirt  Erkek      17.0       Kadın A    https://cdn.dsmcdn.com//ty10/product/media/images/20200807/18/7463388/79269046/1/1_org_zoom.jpg\n",
            "11      104                5        338745787          44743097      7651          604       T-Shirt  Erkek       8.0       Kadın A    https://cdn.dsmcdn.com//ty10/product/media/images/20200828/16/9023785/82976501/1/1_org_zoom.jpg\n",
            "(358913, 11)\n",
            "<class 'pandas.core.frame.DataFrame'>\n",
            "Int64Index: 358913 entries, 7 to 508227\n",
            "Data columns (total 11 columns):\n",
            " #   Column            Non-Null Count   Dtype  \n",
            "---  ------            --------------   -----  \n",
            " 0   user_id           358913 non-null  int64  \n",
            " 1   orderparentid_x   358913 non-null  int64  \n",
            " 2   orderparentid_y   358913 non-null  int64  \n",
            " 3   productcontentid  358913 non-null  int64  \n",
            " 4   brand_id          358913 non-null  int64  \n",
            " 5   category_id       358913 non-null  int64  \n",
            " 6   category_name     358913 non-null  object \n",
            " 7   gender            339456 non-null  object \n",
            " 8   color_id          272164 non-null  float64\n",
            " 9   business_unit     358913 non-null  object \n",
            " 10  ImageLink         358913 non-null  object \n",
            "dtypes: float64(1), int64(6), object(4)\n",
            "memory usage: 32.9+ MB\n",
            "None\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "final_data1 = final_data.iloc[:30000]"
      ],
      "metadata": {
        "id": "cn3tTd6kfK9A"
      },
      "execution_count": 8,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "To create a product recommendation system using a collaborative filtering algorithm, the data needs to be structured so that each record includes information on which items each customer has purchased. Currently, the data consists of individual items purchased by customers. To address this, we will transform the data into a user-to-item matrix, where each row represents a customer and the columns correspond to different products. This format will allow us to use the collaborative filtering algorithm to build the recommendation system."
      ],
      "metadata": {
        "id": "GVjKk8ILDJh0"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "customer_item_matrix = final_data1.pivot_table(\n",
        "    index='user_id',\n",
        "    columns='productcontentid',\n",
        "    values='category_id',\n",
        "    aggfunc='sum'\n",
        ")\n",
        "\n",
        "print(customer_item_matrix.head())"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "MgiNTq3XfK6m",
        "outputId": "d9030e43-12eb-4315-f8f8-56056ca60f52"
      },
      "execution_count": 9,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "productcontentid  52015     52046     52071     52098     52099     52204     \\\n",
            "user_id                                                                        \n",
            "104                    NaN       NaN       NaN       NaN       NaN       NaN   \n",
            "105                    NaN       NaN       NaN       NaN       NaN       NaN   \n",
            "167                    NaN       NaN       NaN       NaN       NaN       NaN   \n",
            "174                    NaN       NaN       NaN       NaN       NaN       NaN   \n",
            "190                    NaN       NaN       NaN       NaN       NaN       NaN   \n",
            "\n",
            "productcontentid  52304     52307     52385     52386     ...  46837954  \\\n",
            "user_id                                                   ...             \n",
            "104                    NaN       NaN       NaN       NaN  ...       NaN   \n",
            "105                    NaN       NaN       NaN       NaN  ...       NaN   \n",
            "167                    NaN       NaN       NaN       NaN  ...       NaN   \n",
            "174                    NaN       NaN       NaN       NaN  ...       NaN   \n",
            "190                    NaN       NaN       NaN       NaN  ...       NaN   \n",
            "\n",
            "productcontentid  46846439  46853373  46855144  46859761  46899769  46928384  \\\n",
            "user_id                                                                        \n",
            "104                    NaN       NaN       NaN       NaN       NaN       NaN   \n",
            "105                    NaN       NaN       NaN       NaN       NaN       NaN   \n",
            "167                    NaN       NaN       NaN       NaN       NaN       NaN   \n",
            "174                  604.0       NaN       NaN       NaN       NaN       NaN   \n",
            "190                    NaN       NaN       NaN       NaN       NaN       NaN   \n",
            "\n",
            "productcontentid  46938926  46958108  47048393  \n",
            "user_id                                         \n",
            "104                    NaN       NaN       NaN  \n",
            "105                    NaN       NaN       NaN  \n",
            "167                    NaN       NaN       NaN  \n",
            "174                    NaN       NaN       NaN  \n",
            "190                    NaN       NaN       NaN  \n",
            "\n",
            "[5 rows x 21558 columns]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "To use this data to build a product recommendation system, we need to convert it into a matrix where each value represents whether or not a particular product was purchased by a particular customer. We can do this by coding the data as 0s and 1s, where a value of 1 indicates that the product was bought by the customer, and a value of 0 indicates that the product was not bought by the customer.\n",
        "\n",
        "For example, the matrix we have now shows the total quantities purchased for each product for each customer. We can transform this into a 0-1 matrix by replacing the quantities with 1s if the product was purchased by the customer, and 0s if the product was not purchased.\n",
        "\n",
        "This will allow us to use the 0-1 matrix as input for a collaborative filtering algorithm to build a product recommendation system."
      ],
      "metadata": {
        "id": "WBT3BL9uDc8r"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "customer_item_matrix = customer_item_matrix.applymap(lambda x: 1 if x > 0 else 0)"
      ],
      "metadata": {
        "id": "po1BktGqfK31"
      },
      "execution_count": 10,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "To transform the data into a 0-1 matrix, we can use the applymap() function from the pandas library. This function applies a given function to each element of a DataFrame.\n",
        "\n",
        "In this case, we are using a lambda function that codes all elements whose values are greater than 0 with 1, and the rest with 0."
      ],
      "metadata": {
        "id": "-Be4rBfGEHYP"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "customer_item_matrix"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 455
        },
        "id": "ZS4GfA5bfK1S",
        "outputId": "fdc99b33-47ef-4827-8a0f-4c44018b2c54"
      },
      "execution_count": 11,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "productcontentid  52015     52046     52071     52098     52099     52204     \\\n",
              "user_id                                                                        \n",
              "104                      0         0         0         0         0         0   \n",
              "105                      0         0         0         0         0         0   \n",
              "167                      0         0         0         0         0         0   \n",
              "174                      0         0         0         0         0         0   \n",
              "190                      0         0         0         0         0         0   \n",
              "...                    ...       ...       ...       ...       ...       ...   \n",
              "20044                    0         0         0         0         0         0   \n",
              "20048                    0         0         0         0         0         0   \n",
              "20050                    0         0         0         0         0         0   \n",
              "20051                    0         0         0         0         0         0   \n",
              "20052                    0         0         0         0         0         0   \n",
              "\n",
              "productcontentid  52304     52307     52385     52386     ...  46837954  \\\n",
              "user_id                                                   ...             \n",
              "104                      0         0         0         0  ...         0   \n",
              "105                      0         0         0         0  ...         0   \n",
              "167                      0         0         0         0  ...         0   \n",
              "174                      0         0         0         0  ...         0   \n",
              "190                      0         0         0         0  ...         0   \n",
              "...                    ...       ...       ...       ...  ...       ...   \n",
              "20044                    0         0         0         0  ...         0   \n",
              "20048                    0         0         0         0  ...         0   \n",
              "20050                    0         0         0         0  ...         0   \n",
              "20051                    0         0         0         0  ...         0   \n",
              "20052                    0         0         0         0  ...         0   \n",
              "\n",
              "productcontentid  46846439  46853373  46855144  46859761  46899769  46928384  \\\n",
              "user_id                                                                        \n",
              "104                      0         0         0         0         0         0   \n",
              "105                      0         0         0         0         0         0   \n",
              "167                      0         0         0         0         0         0   \n",
              "174                      1         0         0         0         0         0   \n",
              "190                      0         0         0         0         0         0   \n",
              "...                    ...       ...       ...       ...       ...       ...   \n",
              "20044                    0         0         0         0         0         0   \n",
              "20048                    0         0         0         0         1         0   \n",
              "20050                    0         0         0         0         0         0   \n",
              "20051                    0         0         0         0         0         0   \n",
              "20052                    0         0         0         0         0         0   \n",
              "\n",
              "productcontentid  46938926  46958108  47048393  \n",
              "user_id                                         \n",
              "104                      0         0         0  \n",
              "105                      0         0         0  \n",
              "167                      0         0         0  \n",
              "174                      0         0         0  \n",
              "190                      0         0         0  \n",
              "...                    ...       ...       ...  \n",
              "20044                    0         0         0  \n",
              "20048                    0         0         0  \n",
              "20050                    0         0         0  \n",
              "20051                    0         0         0  \n",
              "20052                    0         0         0  \n",
              "\n",
              "[3043 rows x 21558 columns]"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-0a3f0328-8032-4b79-a08e-0908e4bb091f\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <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>productcontentid</th>\n",
              "      <th>52015</th>\n",
              "      <th>52046</th>\n",
              "      <th>52071</th>\n",
              "      <th>52098</th>\n",
              "      <th>52099</th>\n",
              "      <th>52204</th>\n",
              "      <th>52304</th>\n",
              "      <th>52307</th>\n",
              "      <th>52385</th>\n",
              "      <th>52386</th>\n",
              "      <th>...</th>\n",
              "      <th>46837954</th>\n",
              "      <th>46846439</th>\n",
              "      <th>46853373</th>\n",
              "      <th>46855144</th>\n",
              "      <th>46859761</th>\n",
              "      <th>46899769</th>\n",
              "      <th>46928384</th>\n",
              "      <th>46938926</th>\n",
              "      <th>46958108</th>\n",
              "      <th>47048393</th>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>user_id</th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>104</th>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>105</th>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>167</th>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>174</th>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>190</th>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>20044</th>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>20048</th>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>20050</th>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>20051</th>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>20052</th>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>3043 rows × 21558 columns</p>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-0a3f0328-8032-4b79-a08e-0908e4bb091f')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-0a3f0328-8032-4b79-a08e-0908e4bb091f button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-0a3f0328-8032-4b79-a08e-0908e4bb091f');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 11
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Now we have a 0-1 matrix that we can use as input for a collaborative filtering algorithm to build a product recommendation system."
      ],
      "metadata": {
        "id": "oJtHboZxD1uJ"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "# **Collaborative filtering**"
      ],
      "metadata": {
        "id": "d2Ql3PP7EBDi"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "There are two approaches to building a product recommendation system: the user-based approach and the item-based approach.\n",
        "\n",
        "In the user-based approach, the similarities between users are calculated based on their item purchase history. This means that users who have purchased similar items are considered more similar to each other.\n",
        "\n",
        "In the item-based approach, on the other hand, the similarities between items are calculated based on which items are often purchased together. This means that items that are frequently purchased together are considered more similar to each other.\n",
        "\n",
        "Both of these approaches can be used to build a product recommendation system, and which one is more appropriate will depend on the specific needs of the application.\n",
        "\n",
        "To measure the similarity between users or between articles, we will use the cosine_similarity method in the scikit-learn package."
      ],
      "metadata": {
        "id": "GjEgbWImEmVv"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.metrics.pairwise import cosine_similarity"
      ],
      "metadata": {
        "id": "8v2ZfWdVfKy6"
      },
      "execution_count": 12,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "# **User-based collaboration filters and recommendations**"
      ],
      "metadata": {
        "id": "ixKRfFDqEvme"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "To build a user-based collaborative filtering algorithm, it is necessary to calculate the cosine similarities between users. Cosine similarity is a measure of similarity between two non-zero vectors of an inner product space that measures the cosine of the angle between them. In the context of a recommendation system, the vectors represent the item purchase histories of different users, and the cosine similarity measures the similarity between two users based on the items they have purchased.\n",
        "\n",
        "To calculate the cosine similarities between users, we can use the cosine_similarity() function from the sklearn.metrics.pairwise module. This function takes as input a matrix of user vectors and returns a matrix of cosine similarities between the users."
      ],
      "metadata": {
        "id": "BirvF7LTFDaI"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "user_user_sim_matrix = pd.DataFrame(\n",
        "    cosine_similarity(customer_item_matrix)\n",
        ")"
      ],
      "metadata": {
        "id": "WyENoMqffKwP"
      },
      "execution_count": 13,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "user_user_sim_matrix "
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 424
        },
        "id": "Tz0yz1CnfKt9",
        "outputId": "f54fa8c1-b3ed-4f02-8a5f-c7011c8581a6"
      },
      "execution_count": 14,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "      0     1     2     3     4     5     6     7     8     9     ...  3033  \\\n",
              "0      1.0   0.0   0.0   0.0   0.0   0.0   0.0   0.0   0.0   0.0  ...   0.0   \n",
              "1      0.0   1.0   0.0   0.0   0.0   0.0   0.0   0.0   0.0   0.0  ...   0.0   \n",
              "2      0.0   0.0   1.0   0.0   0.0   0.0   0.0   0.0   0.0   0.0  ...   0.0   \n",
              "3      0.0   0.0   0.0   1.0   0.0   0.0   0.0   0.0   0.0   0.0  ...   0.0   \n",
              "4      0.0   0.0   0.0   0.0   1.0   0.0   0.0   0.0   0.0   0.0  ...   0.0   \n",
              "...    ...   ...   ...   ...   ...   ...   ...   ...   ...   ...  ...   ...   \n",
              "3038   0.0   0.0   0.0   0.0   0.0   0.0   0.0   0.0   0.0   0.0  ...   0.0   \n",
              "3039   0.0   0.0   0.0   0.0   0.0   0.0   0.0   0.0   0.0   0.0  ...   0.0   \n",
              "3040   0.0   0.0   0.0   0.0   0.0   0.0   0.0   0.0   0.0   0.0  ...   0.0   \n",
              "3041   0.0   0.0   0.0   0.0   0.0   0.0   0.0   0.0   0.0   0.0  ...   0.0   \n",
              "3042   0.0   0.0   0.0   0.0   0.0   0.0   0.0   0.0   0.0   0.0  ...   0.0   \n",
              "\n",
              "      3034  3035  3036  3037  3038  3039  3040  3041  3042  \n",
              "0      0.0   0.0   0.0   0.0   0.0   0.0   0.0   0.0   0.0  \n",
              "1      0.0   0.0   0.0   0.0   0.0   0.0   0.0   0.0   0.0  \n",
              "2      0.0   0.0   0.0   0.0   0.0   0.0   0.0   0.0   0.0  \n",
              "3      0.0   0.0   0.0   0.0   0.0   0.0   0.0   0.0   0.0  \n",
              "4      0.0   0.0   0.0   0.0   0.0   0.0   0.0   0.0   0.0  \n",
              "...    ...   ...   ...   ...   ...   ...   ...   ...   ...  \n",
              "3038   0.0   0.0   0.0   0.0   1.0   0.0   0.0   0.0   0.0  \n",
              "3039   0.0   0.0   0.0   0.0   0.0   1.0   0.0   0.0   0.0  \n",
              "3040   0.0   0.0   0.0   0.0   0.0   0.0   1.0   0.0   0.0  \n",
              "3041   0.0   0.0   0.0   0.0   0.0   0.0   0.0   1.0   0.0  \n",
              "3042   0.0   0.0   0.0   0.0   0.0   0.0   0.0   0.0   1.0  \n",
              "\n",
              "[3043 rows x 3043 columns]"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-75d0a67c-d615-44f6-919e-0f037f5248ca\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <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>0</th>\n",
              "      <th>1</th>\n",
              "      <th>2</th>\n",
              "      <th>3</th>\n",
              "      <th>4</th>\n",
              "      <th>5</th>\n",
              "      <th>6</th>\n",
              "      <th>7</th>\n",
              "      <th>8</th>\n",
              "      <th>9</th>\n",
              "      <th>...</th>\n",
              "      <th>3033</th>\n",
              "      <th>3034</th>\n",
              "      <th>3035</th>\n",
              "      <th>3036</th>\n",
              "      <th>3037</th>\n",
              "      <th>3038</th>\n",
              "      <th>3039</th>\n",
              "      <th>3040</th>\n",
              "      <th>3041</th>\n",
              "      <th>3042</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>0.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3038</th>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3039</th>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3040</th>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3041</th>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3042</th>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1.0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>3043 rows × 3043 columns</p>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-75d0a67c-d615-44f6-919e-0f037f5248ca')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-75d0a67c-d615-44f6-919e-0f037f5248ca button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-75d0a67c-d615-44f6-919e-0f037f5248ca');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 14
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "In this code, we are using the cosine_similarity() function from the sklearn.metrics.pairwise module to calculate the cosine similarities between the samples. This function returns the similarities as an array.\n",
        "\n",
        "Then, we create a DataFrame from this output array and store it in a variable called user_user_sim_matrix, which represents the user-user similarity array. The index and column names of the DataFrame may not be easily interpretable, as each column and row represents an individual customer.\n",
        "\n",
        "To make the DataFrame more understandable, we can consider renaming the index and column names to represent the customer ID's which is user_id. This will make it easier to understand which customers are being compared in the DataFrame."
      ],
      "metadata": {
        "id": "1QsgnlO2Fk9H"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "user_user_sim_matrix.columns = customer_item_matrix.index"
      ],
      "metadata": {
        "id": "UXfVC4RHfKrm"
      },
      "execution_count": 15,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "user_user_sim_matrix"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 424
        },
        "id": "ov7nqdxVfKpR",
        "outputId": "e137612c-7868-4126-9670-15cb11e5bfa6"
      },
      "execution_count": 16,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "user_id  104    105    167    174    190    197    216    222    232    234    \\\n",
              "0          1.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0   \n",
              "1          0.0    1.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0   \n",
              "2          0.0    0.0    1.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0   \n",
              "3          0.0    0.0    0.0    1.0    0.0    0.0    0.0    0.0    0.0    0.0   \n",
              "4          0.0    0.0    0.0    0.0    1.0    0.0    0.0    0.0    0.0    0.0   \n",
              "...        ...    ...    ...    ...    ...    ...    ...    ...    ...    ...   \n",
              "3038       0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0   \n",
              "3039       0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0   \n",
              "3040       0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0   \n",
              "3041       0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0   \n",
              "3042       0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0   \n",
              "\n",
              "user_id  ...  20035  20037  20038  20039  20042  20044  20048  20050  20051  \\\n",
              "0        ...    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0   \n",
              "1        ...    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0   \n",
              "2        ...    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0   \n",
              "3        ...    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0   \n",
              "4        ...    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0   \n",
              "...      ...    ...    ...    ...    ...    ...    ...    ...    ...    ...   \n",
              "3038     ...    0.0    0.0    0.0    0.0    0.0    1.0    0.0    0.0    0.0   \n",
              "3039     ...    0.0    0.0    0.0    0.0    0.0    0.0    1.0    0.0    0.0   \n",
              "3040     ...    0.0    0.0    0.0    0.0    0.0    0.0    0.0    1.0    0.0   \n",
              "3041     ...    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0    1.0   \n",
              "3042     ...    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0   \n",
              "\n",
              "user_id  20052  \n",
              "0          0.0  \n",
              "1          0.0  \n",
              "2          0.0  \n",
              "3          0.0  \n",
              "4          0.0  \n",
              "...        ...  \n",
              "3038       0.0  \n",
              "3039       0.0  \n",
              "3040       0.0  \n",
              "3041       0.0  \n",
              "3042       1.0  \n",
              "\n",
              "[3043 rows x 3043 columns]"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-ff4df88e-4fe7-43d3-84db-02733d3d06e6\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <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>user_id</th>\n",
              "      <th>104</th>\n",
              "      <th>105</th>\n",
              "      <th>167</th>\n",
              "      <th>174</th>\n",
              "      <th>190</th>\n",
              "      <th>197</th>\n",
              "      <th>216</th>\n",
              "      <th>222</th>\n",
              "      <th>232</th>\n",
              "      <th>234</th>\n",
              "      <th>...</th>\n",
              "      <th>20035</th>\n",
              "      <th>20037</th>\n",
              "      <th>20038</th>\n",
              "      <th>20039</th>\n",
              "      <th>20042</th>\n",
              "      <th>20044</th>\n",
              "      <th>20048</th>\n",
              "      <th>20050</th>\n",
              "      <th>20051</th>\n",
              "      <th>20052</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>0.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3038</th>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3039</th>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3040</th>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3041</th>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3042</th>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1.0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>3043 rows × 3043 columns</p>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-ff4df88e-4fe7-43d3-84db-02733d3d06e6')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-ff4df88e-4fe7-43d3-84db-02733d3d06e6 button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-ff4df88e-4fe7-43d3-84db-02733d3d06e6');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 16
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "The user similarity matrix shows the cosine similarities between pairs of users. As expected, the cosine similarity between a user and itself is 1, which is reflected in the diagonal elements of the matrix.\n",
        "\n",
        "The other elements in the matrix represent the cosine similarity between two different users. For example, if the measure of cosine similarity between users 1031 and 822 is 0.2857, it means that these two users have purchased similar items and can be considered similar to each other.\n",
        "\n",
        "By examining the user similarity matrix, it is possible to see which users are similar to each other and which users have purchased similar items. This information can be used to build a user-based collaborative filtering algorithm for making product recommendations."
      ],
      "metadata": {
        "id": "g3fQ96W6F6OT"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "user_user_sim_matrix['user_id'] = customer_item_matrix.index\n",
        "user_user_sim_matrix = user_user_sim_matrix.set_index('user_id')"
      ],
      "metadata": {
        "id": "O_W91NEafKmg"
      },
      "execution_count": 17,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "user_user_sim_matrix.loc[1031.0].sort_values(ascending=False)\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "qVw7sIb5fKkD",
        "outputId": "b2f0699c-ef9e-490a-fb3d-48452ec9b737"
      },
      "execution_count": 18,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "user_id\n",
              "1031     1.000000\n",
              "822      0.285714\n",
              "11004    0.285714\n",
              "3057     0.267261\n",
              "12860    0.169031\n",
              "           ...   \n",
              "9900     0.000000\n",
              "9902     0.000000\n",
              "9911     0.000000\n",
              "9912     0.000000\n",
              "20052    0.000000\n",
              "Name: 1031, Length: 3043, dtype: float64"
            ]
          },
          "metadata": {},
          "execution_count": 18
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "items_bought_by_A = set(customer_item_matrix.loc[1031.0].iloc[\n",
        "    customer_item_matrix.loc[1031.0].to_numpy().nonzero()\n",
        "].index)"
      ],
      "metadata": {
        "id": "lOBAcasHfKhq"
      },
      "execution_count": 19,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "To build a recommendation system using a user-based collaborative filtering algorithm, we can follow the strategy outlined in your description.\n",
        "\n",
        "\\\n",
        "\n",
        "1. Identify the items that users 1031 and 822 have already purchased. This can be done by looking at the item purchase history data for these users.\n",
        "2. Find the items that user 822 has not purchased, but user 1031 has. These are the items that we will recommend to user 822.\n",
        "3. Assume that user 822 has a high probability of buying these items because they have purchased similar items in the past, according to the user similarity matrix.\n",
        "4. Recommend the list of items to user 822. \n",
        "\n",
        "\\\n",
        "\n",
        "By following these steps, we can use the user similarity matrix and the item purchase history data to make personalized recommendations to user 822 based on the items that similar users (in this case, user 1031) have purchased.\n"
      ],
      "metadata": {
        "id": "4PRREjczGRp4"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "items_bought_by_A"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "cJPDtHmhjJlR",
        "outputId": "a6184c42-f07f-42b6-acfd-c5dccd9abc99"
      },
      "execution_count": 20,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "{34145548, 35740100, 36353009, 36709507, 37187972, 38922211, 42692453}"
            ]
          },
          "metadata": {},
          "execution_count": 20
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "items_bought_by_B = set(customer_item_matrix.loc[822.0].iloc[\n",
        "    customer_item_matrix.loc[517].to_numpy().nonzero()\n",
        "].index)"
      ],
      "metadata": {
        "id": "zh_-tG6AjJiu"
      },
      "execution_count": 21,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "items_bought_by_B"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "_Lg5shmxjJgN",
        "outputId": "54b82f6f-1459-4701-ed14-02ecaeca7e27"
      },
      "execution_count": 22,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "{39812560, 39832331, 43524099, 44606911}"
            ]
          },
          "metadata": {},
          "execution_count": 22
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "items_to_recommend_to_B = items_bought_by_A - items_bought_by_B\n",
        "items_to_recommend_to_B"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ZG_2veBPjJd0",
        "outputId": "df1a1db6-d625-41b2-a6d4-d0aa84d17116"
      },
      "execution_count": 23,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "{34145548, 35740100, 36353009, 36709507, 37187972, 38922211, 42692453}"
            ]
          },
          "metadata": {},
          "execution_count": 23
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "final_data1.loc[\n",
        "    final_data1['productcontentid'].isin(items_to_recommend_to_B), \n",
        "    ['productcontentid', 'category_name','gender','ImageLink']\n",
        "].drop_duplicates().set_index('productcontentid')"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 300
        },
        "id": "XVOFGB2hjJbc",
        "outputId": "4c9905bc-1274-48bc-a95f-615d5ebbe06d"
      },
      "execution_count": 24,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "                      category_name gender  \\\n",
              "productcontentid                             \n",
              "36353009                    T-Shirt  Erkek   \n",
              "38922211          Polo Yaka T-shirt  Erkek   \n",
              "42692453                    T-Shirt  Erkek   \n",
              "37187972                    T-Shirt  Erkek   \n",
              "36709507                    T-Shirt  Erkek   \n",
              "35740100                    T-Shirt  Erkek   \n",
              "34145548                    T-Shirt  Erkek   \n",
              "\n",
              "                                                          ImageLink  \n",
              "productcontentid                                                     \n",
              "36353009          https://cdn.dsmcdn.com//assets/product/media/i...  \n",
              "38922211          https://cdn.dsmcdn.com//ty1/product/media/imag...  \n",
              "42692453          https://cdn.dsmcdn.com//ty7/product/media/imag...  \n",
              "37187972          https://cdn.dsmcdn.com//ty1/product/media/imag...  \n",
              "36709507          https://cdn.dsmcdn.com//ty1/product/media/imag...  \n",
              "35740100          https://cdn.dsmcdn.com//assets/product/media/i...  \n",
              "34145548          https://cdn.dsmcdn.com//assets/product/media/i...  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-9a94d7c1-5836-4f9c-93b3-4dc793d49eb2\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <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>category_name</th>\n",
              "      <th>gender</th>\n",
              "      <th>ImageLink</th>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>productcontentid</th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>36353009</th>\n",
              "      <td>T-Shirt</td>\n",
              "      <td>Erkek</td>\n",
              "      <td>https://cdn.dsmcdn.com//assets/product/media/i...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>38922211</th>\n",
              "      <td>Polo Yaka T-shirt</td>\n",
              "      <td>Erkek</td>\n",
              "      <td>https://cdn.dsmcdn.com//ty1/product/media/imag...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>42692453</th>\n",
              "      <td>T-Shirt</td>\n",
              "      <td>Erkek</td>\n",
              "      <td>https://cdn.dsmcdn.com//ty7/product/media/imag...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>37187972</th>\n",
              "      <td>T-Shirt</td>\n",
              "      <td>Erkek</td>\n",
              "      <td>https://cdn.dsmcdn.com//ty1/product/media/imag...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>36709507</th>\n",
              "      <td>T-Shirt</td>\n",
              "      <td>Erkek</td>\n",
              "      <td>https://cdn.dsmcdn.com//ty1/product/media/imag...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>35740100</th>\n",
              "      <td>T-Shirt</td>\n",
              "      <td>Erkek</td>\n",
              "      <td>https://cdn.dsmcdn.com//assets/product/media/i...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>34145548</th>\n",
              "      <td>T-Shirt</td>\n",
              "      <td>Erkek</td>\n",
              "      <td>https://cdn.dsmcdn.com//assets/product/media/i...</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-9a94d7c1-5836-4f9c-93b3-4dc793d49eb2')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-9a94d7c1-5836-4f9c-93b3-4dc793d49eb2 button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-9a94d7c1-5836-4f9c-93b3-4dc793d49eb2');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 24
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "User-based collaborative filtering is a technique that can be used to make personalized product recommendations to individual customers based on their purchase history and the purchase histories of similar users. By recommending products that are likely to be of interest to each target customer, you can increase conversions and improve the effectiveness of your marketing efforts.\n",
        "\n",
        "\n",
        "However, user-based collaborative filtering has one major disadvantage: it relies on the individual customer's purchase history, which means that it is not suitable for making recommendations to new customers who have no purchase history. In this case, there is not enough data to compare the new customer to other users, and the recommendation system will not be able to make accurate recommendations.\n",
        "\n",
        "To handle this problem, we can use item-based collaborative filtering instead. In item-based collaborative filtering, the similarities between items are calculated based on which items are often purchased together. This means that items that are frequently purchased together are considered more similar to each other.\n",
        "\n",
        "We can use item-based collaborative filtering to make recommendations to new customers, as well as to existing customers, by considering the items that they have purchased and recommending similar items. We will discuss item-based collaborative filtering in more detail in the next section.\n",
        "\n",
        "\n",
        "\n",
        "\n"
      ],
      "metadata": {
        "id": "VEYz6MJnHcOL"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "# **Collaborative filtering based on articles and recommendations**"
      ],
      "metadata": {
        "id": "reAx8MS8HpKQ"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "Item-based collaborative filtering is a technique for making product recommendations that is similar to user-based collaborative filtering, except that it uses measures of similarity between items rather than between users.\n",
        "\n",
        "To implement item-based collaborative filtering, we need to calculate the cosine similarities between pairs of items. Cosine similarity is a measure of similarity between two non-zero vectors of an inner product space that measures the cosine of the angle between them. In the context of a recommendation system, the vectors represent the item purchase histories of different users, and the cosine similarity measures the similarity between two items based on how often they are purchased together by different users.\n",
        "\n",
        "Once we have the cosine similarities between the items, we can use them to build an item-based collaborative filtering algorithm for making product recommendations."
      ],
      "metadata": {
        "id": "kIzFi9lSH_4B"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "item_item_sim_matrix = pd.DataFrame(\n",
        "    cosine_similarity(customer_item_matrix.T)\n",
        ")"
      ],
      "metadata": {
        "id": "FzjVvSMUjJWe"
      },
      "execution_count": 25,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "The process of calculating the cosine similarities between items using item-based collaborative filtering is similar to calculating the cosine similarities between users using user-based collaborative filtering. The main difference is that we need to transpose the customer-item matrix so that the rows represent individual items and the columns represent the customers.\n",
        "\n",
        "To use the cosine_similarity function from the metrics.pairwise module in the sklearn package. To correctly name the indexes and columns with the product codes"
      ],
      "metadata": {
        "id": "7QsMYGLdIZ5c"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "item_item_sim_matrix"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 424
        },
        "id": "zsA3AY0Kjfb4",
        "outputId": "b4fa92c6-a0be-4485-dee5-e0243409424f"
      },
      "execution_count": 26,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "       0      1      2      3      4      5      6      7      8      9      \\\n",
              "0        1.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0   \n",
              "1        0.0    1.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0   \n",
              "2        0.0    0.0    1.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0   \n",
              "3        0.0    0.0    0.0    1.0    0.0    0.0    0.0    0.0    0.0    0.0   \n",
              "4        0.0    0.0    0.0    0.0    1.0    0.0    1.0    0.0    0.0    0.0   \n",
              "...      ...    ...    ...    ...    ...    ...    ...    ...    ...    ...   \n",
              "21553    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0   \n",
              "21554    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0   \n",
              "21555    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0   \n",
              "21556    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0   \n",
              "21557    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0   \n",
              "\n",
              "       ...  21548  21549  21550  21551  21552  21553  21554  21555  21556  \\\n",
              "0      ...    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0   \n",
              "1      ...    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0   \n",
              "2      ...    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0   \n",
              "3      ...    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0   \n",
              "4      ...    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0   \n",
              "...    ...    ...    ...    ...    ...    ...    ...    ...    ...    ...   \n",
              "21553  ...    0.0    0.0    0.0    0.0    0.0    1.0    0.0    0.0    0.0   \n",
              "21554  ...    0.0    0.0    0.0    0.0    0.0    0.0    1.0    0.0    0.0   \n",
              "21555  ...    0.0    0.0    0.0    0.0    0.0    0.0    0.0    1.0    0.0   \n",
              "21556  ...    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0    1.0   \n",
              "21557  ...    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0    0.0   \n",
              "\n",
              "       21557  \n",
              "0        0.0  \n",
              "1        0.0  \n",
              "2        0.0  \n",
              "3        0.0  \n",
              "4        0.0  \n",
              "...      ...  \n",
              "21553    0.0  \n",
              "21554    0.0  \n",
              "21555    0.0  \n",
              "21556    0.0  \n",
              "21557    1.0  \n",
              "\n",
              "[21558 rows x 21558 columns]"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-4f459606-c455-4dfb-9e22-6099f381b8ef\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <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>0</th>\n",
              "      <th>1</th>\n",
              "      <th>2</th>\n",
              "      <th>3</th>\n",
              "      <th>4</th>\n",
              "      <th>5</th>\n",
              "      <th>6</th>\n",
              "      <th>7</th>\n",
              "      <th>8</th>\n",
              "      <th>9</th>\n",
              "      <th>...</th>\n",
              "      <th>21548</th>\n",
              "      <th>21549</th>\n",
              "      <th>21550</th>\n",
              "      <th>21551</th>\n",
              "      <th>21552</th>\n",
              "      <th>21553</th>\n",
              "      <th>21554</th>\n",
              "      <th>21555</th>\n",
              "      <th>21556</th>\n",
              "      <th>21557</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>0.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>21553</th>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>21554</th>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>21555</th>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>21556</th>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>21557</th>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1.0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>21558 rows × 21558 columns</p>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-4f459606-c455-4dfb-9e22-6099f381b8ef')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-4f459606-c455-4dfb-9e22-6099f381b8ef button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-4f459606-c455-4dfb-9e22-6099f381b8ef');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 26
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "item_item_sim_matrix.columns = customer_item_matrix.T.index\n",
        "\n",
        "item_item_sim_matrix['productcontentid'] = customer_item_matrix.T.index\n",
        "item_item_sim_matrix = item_item_sim_matrix.set_index('productcontentid')"
      ],
      "metadata": {
        "id": "2BMzEljgjfW4"
      },
      "execution_count": 30,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "top_5_similar_items = list(\n",
        "   item_item_sim_matrix.loc[52098.0].sort_values(ascending=False).iloc[:6].index\n",
        ")"
      ],
      "metadata": {
        "id": "yBciuP5rjfUi"
      },
      "execution_count": 31,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "As before, diagonal elements have values of 1. This is because the similarity between an element and itself is 1, which means that the two are identical. The rest of the elements contain the values for measuring the similarity between the elements based on the calculation of cosine similarity.\n",
        "\n",
        "For example, if you look at the similarity matrix between elements above, the cosine similarity between the element with StockCode 52098 and the element with StockCode 39646446 is 0.707107. \n",
        "\n",
        "On the other hand, the cosine similarity between item 52098 and item 321912 is 0.288675 This suggests that the item with StockCode 4443412 is more similar to the item with StockCode 52098 than the item with StockCode 10125 to the item with StockCode 52098.\n",
        "\n",
        "The strategy for making the product recommendation using this item similarity matrix is similar to the one we did using the user-based approach.\n",
        "\n",
        "First, for the product given that the target customer bought, we are going to find the most similar items from the item-to-item similarity matrix that we just built.\n",
        "Then, we're going to recommend these similar items to the customer because those similar items were bought by other customers who have purchased the product that the target customer initially bought."
      ],
      "metadata": {
        "id": "6kbQz-H3I29M"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "item_item_sim_matrix[52098].sort_values()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "7IJLyxkA-VkG",
        "outputId": "bfe5a180-6fd5-4f02-bc53-7e1e708060c6"
      },
      "execution_count": 32,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "productcontentid\n",
              "52015       0.000000\n",
              "38354263    0.000000\n",
              "38353503    0.000000\n",
              "38346878    0.000000\n",
              "38346857    0.000000\n",
              "              ...   \n",
              "4443412     0.707107\n",
              "43360636    0.707107\n",
              "953672      0.707107\n",
              "39646446    0.707107\n",
              "52098       1.000000\n",
              "Name: 52098, Length: 21558, dtype: float64"
            ]
          },
          "metadata": {},
          "execution_count": 32
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "To make product recommendations using an item-based collaborative filtering algorithm, we can follow a strategy similar to the one used in user-based collaborative filtering.\n",
        "\n",
        "\\\n",
        "\n",
        "The steps involved in this process are as follows:\n",
        "\n",
        "1. Identify the item that the target customer has purchased.\n",
        "2. Find the most similar items to this item using the item similarity matrix. These are the items that we will recommend to the customer.\n",
        "4. Recommend the similar items to the customer because they have been purchased by other customers who have also purchased the target item.\n",
        " \n",
        " \n",
        " For example, suppose we have a new customer who has just purchased a product with productcontentid xxx, and we want to include in our marketing emails some products that this customer is likely to buy. To find the items most similar to the item with productcontentid."
      ],
      "metadata": {
        "id": "5E5Lmg3GNQ1g"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "final_data.loc[\n",
        "    final_data['productcontentid'].isin(top_5_similar_items), \n",
        "    ['productcontentid', 'category_name', 'ImageLink']\n",
        "].drop_duplicates().set_index('productcontentid').loc[top_5_similar_items]"
      ],
      "metadata": {
        "id": "aOl3puqJjfSC",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 269
        },
        "outputId": "ab3e6f58-e09c-41f1-890e-6767d4969737"
      },
      "execution_count": 33,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "                        category_name  \\\n",
              "productcontentid                        \n",
              "52098                         El Yayı   \n",
              "40488929                        Havuz   \n",
              "34420340          Fitness - Kondisyon   \n",
              "39216939                        Forma   \n",
              "952415            Fitness - Kondisyon   \n",
              "953672              Egzersiz Aletleri   \n",
              "\n",
              "                                                          ImageLink  \n",
              "productcontentid                                                     \n",
              "52098             https://cdn.dsmcdn.com//assets/product/media/i...  \n",
              "40488929          https://cdn.dsmcdn.com//ty3/product/media/imag...  \n",
              "34420340          https://cdn.dsmcdn.com//assets/product/media/i...  \n",
              "39216939          https://cdn.dsmcdn.com//ty1/product/media/imag...  \n",
              "952415            https://cdn.dsmcdn.com//ty3/product/media/imag...  \n",
              "953672            https://cdn.dsmcdn.com//Assets/ProductImages/3...  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-8d87ea63-27bd-4cd0-b32e-27b6e544a209\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <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>category_name</th>\n",
              "      <th>ImageLink</th>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>productcontentid</th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>52098</th>\n",
              "      <td>El Yayı</td>\n",
              "      <td>https://cdn.dsmcdn.com//assets/product/media/i...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>40488929</th>\n",
              "      <td>Havuz</td>\n",
              "      <td>https://cdn.dsmcdn.com//ty3/product/media/imag...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>34420340</th>\n",
              "      <td>Fitness - Kondisyon</td>\n",
              "      <td>https://cdn.dsmcdn.com//assets/product/media/i...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>39216939</th>\n",
              "      <td>Forma</td>\n",
              "      <td>https://cdn.dsmcdn.com//ty1/product/media/imag...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>952415</th>\n",
              "      <td>Fitness - Kondisyon</td>\n",
              "      <td>https://cdn.dsmcdn.com//ty3/product/media/imag...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>953672</th>\n",
              "      <td>Egzersiz Aletleri</td>\n",
              "      <td>https://cdn.dsmcdn.com//Assets/ProductImages/3...</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-8d87ea63-27bd-4cd0-b32e-27b6e544a209')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-8d87ea63-27bd-4cd0-b32e-27b6e544a209 button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-8d87ea63-27bd-4cd0-b32e-27b6e544a209');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 33
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Using an item-based collaborative filtering algorithm, you can make personalized product recommendations for individual customers based on their purchase history and the purchase histories of other customers.\n",
        "\n",
        "For example, suppose you have a customer who has just purchased a \"El Yayı\", and you want to include in your marketing emails some additional product recommendations for this customer. By analyzing the purchase histories of other customers who have also purchased \"El Yayı\", you can identify other items that are frequently purchased along with ceramic \"El Yayı\". These items can then be included in your marketing messages as additional product recommendations for the target customer. \n",
        "\n",
        "In this way, you can use an item-based collaborative filtering algorithm to make personalized product recommendations that are likely to be of interest to your customers, which can result in higher customer conversion rates."
      ],
      "metadata": {
        "id": "5v1HVgEGPHT8"
      }
    }
  ]
}