sports_outdoor_review_W2V.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "sports_outdoor_review_W2V.ipynb",
      "provenance": [],
      "collapsed_sections": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "d4tO8VlK3sHk",
        "outputId": "c5f44de5-f12d-4a9a-a097-a5417d393239"
      },
      "source": [
        "!pip install gensim\n",
        "!pip install python-Levenshtein"
      ],
      "execution_count": 1,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Requirement already satisfied: gensim in /usr/local/lib/python3.7/dist-packages (3.6.0)\n",
            "Requirement already satisfied: six>=1.5.0 in /usr/local/lib/python3.7/dist-packages (from gensim) (1.15.0)\n",
            "Requirement already satisfied: smart-open>=1.2.1 in /usr/local/lib/python3.7/dist-packages (from gensim) (5.0.0)\n",
            "Requirement already satisfied: scipy>=0.18.1 in /usr/local/lib/python3.7/dist-packages (from gensim) (1.4.1)\n",
            "Requirement already satisfied: numpy>=1.11.3 in /usr/local/lib/python3.7/dist-packages (from gensim) (1.19.5)\n",
            "Requirement already satisfied: python-Levenshtein in /usr/local/lib/python3.7/dist-packages (0.12.2)\n",
            "Requirement already satisfied: setuptools in /usr/local/lib/python3.7/dist-packages (from python-Levenshtein) (56.1.0)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "6AQTpOEn5YfL"
      },
      "source": [
        "import gensim\n",
        "import pandas as pd\n",
        "import tensorflow as tf"
      ],
      "execution_count": 2,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Nq3DpszsCJEi"
      },
      "source": [
        "\n",
        "#Reading and Exploring the Dataset\n",
        "\n",
        "The dataset we are using here is a subset of Amazon reviews from the Sports & Outdoors category. The data is stored as a JSON file and can be read using pandas.\n",
        "\n",
        "Link to the Dataset: http://snap.stanford.edu/data/amazon/productGraph/categoryFiles/reviews_Sports_and_Outdoors_5.json.gz"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "BuVU4d_K5e9a"
      },
      "source": [
        "#from google.colab import drive\n",
        "#drive.mount('/content/drive')"
      ],
      "execution_count": 3,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 568
        },
        "id": "RwF5iVV66MTR",
        "outputId": "88498fd1-602e-4681-e442-bf5a62defdf4"
      },
      "source": [
        "df= pd.read_json(\"/content/drive/MyDrive/Colab Notebooks/Sports_and_Outdoors_5.json\", lines=True)\n",
        "df.head()"
      ],
      "execution_count": 4,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>reviewerID</th>\n",
              "      <th>asin</th>\n",
              "      <th>reviewerName</th>\n",
              "      <th>helpful</th>\n",
              "      <th>reviewText</th>\n",
              "      <th>overall</th>\n",
              "      <th>summary</th>\n",
              "      <th>unixReviewTime</th>\n",
              "      <th>reviewTime</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>AIXZKN4ACSKI</td>\n",
              "      <td>1881509818</td>\n",
              "      <td>David Briner</td>\n",
              "      <td>[0, 0]</td>\n",
              "      <td>This came in on time and I am veru happy with ...</td>\n",
              "      <td>5</td>\n",
              "      <td>Woks very good</td>\n",
              "      <td>1390694400</td>\n",
              "      <td>01 26, 2014</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>A1L5P841VIO02V</td>\n",
              "      <td>1881509818</td>\n",
              "      <td>Jason A. Kramer</td>\n",
              "      <td>[1, 1]</td>\n",
              "      <td>I had a factory Glock tool that I was using fo...</td>\n",
              "      <td>5</td>\n",
              "      <td>Works as well as the factory tool</td>\n",
              "      <td>1328140800</td>\n",
              "      <td>02 2, 2012</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>AB2W04NI4OEAD</td>\n",
              "      <td>1881509818</td>\n",
              "      <td>J. Fernald</td>\n",
              "      <td>[2, 2]</td>\n",
              "      <td>If you don't have a 3/32 punch or would like t...</td>\n",
              "      <td>4</td>\n",
              "      <td>It's a punch, that's all.</td>\n",
              "      <td>1330387200</td>\n",
              "      <td>02 28, 2012</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>A148SVSWKTJKU6</td>\n",
              "      <td>1881509818</td>\n",
              "      <td>Jusitn A. Watts \"Maverick9614\"</td>\n",
              "      <td>[0, 0]</td>\n",
              "      <td>This works no better than any 3/32 punch you w...</td>\n",
              "      <td>4</td>\n",
              "      <td>It's a punch with a Glock logo.</td>\n",
              "      <td>1328400000</td>\n",
              "      <td>02 5, 2012</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>AAAWJ6LW9WMOO</td>\n",
              "      <td>1881509818</td>\n",
              "      <td>Material Man</td>\n",
              "      <td>[0, 0]</td>\n",
              "      <td>I purchased this thinking maybe I need a speci...</td>\n",
              "      <td>4</td>\n",
              "      <td>Ok,tool does what a regular punch does.</td>\n",
              "      <td>1366675200</td>\n",
              "      <td>04 23, 2013</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "       reviewerID        asin  ... unixReviewTime   reviewTime\n",
              "0    AIXZKN4ACSKI  1881509818  ...     1390694400  01 26, 2014\n",
              "1  A1L5P841VIO02V  1881509818  ...     1328140800   02 2, 2012\n",
              "2   AB2W04NI4OEAD  1881509818  ...     1330387200  02 28, 2012\n",
              "3  A148SVSWKTJKU6  1881509818  ...     1328400000   02 5, 2012\n",
              "4   AAAWJ6LW9WMOO  1881509818  ...     1366675200  04 23, 2013\n",
              "\n",
              "[5 rows x 9 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 4
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 52
        },
        "id": "q269p4W46Quy",
        "outputId": "79f5c3c7-745f-4895-cbf3-ace16b7cfe8a"
      },
      "source": [
        "df.reviewText[0]"
      ],
      "execution_count": 5,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "string"
            },
            "text/plain": [
              "'This came in on time and I am veru happy with it, I haved used it already and it makes taking out the pins in my glock 32 very easy'"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 5
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "nI6z1Vcp6pjR",
        "outputId": "e6d593dd-ab5e-4171-a8ef-cb27bafe0ba4"
      },
      "source": [
        "df.shape"
      ],
      "execution_count": 6,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(296337, 9)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 6
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "pUI3X_BKCe2C"
      },
      "source": [
        "# Simple Preprocessing & Tokenization\n",
        "The first thing to do for any data science task is to clean the data. For NLP, we apply various processing like converting all the words to lower case, trimming spaces, removing punctuations. This is something we will do over here too.\n",
        "\n",
        "Additionally, we can also remove stop words like 'and', 'or', 'is', 'the', 'a', 'an' and convert words to their root forms like 'running' to 'run'."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "cWDgRuG26u3p",
        "outputId": "4fb6b04a-5165-40fc-da9c-615cf8331749"
      },
      "source": [
        "review_text=df.reviewText.apply(gensim.utils.simple_preprocess)\n",
        "review_text.head()"
      ],
      "execution_count": 7,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0    [this, came, in, on, time, and, am, veru, happ...\n",
              "1    [had, factory, glock, tool, that, was, using, ...\n",
              "2    [if, you, don, have, punch, or, would, like, t...\n",
              "3    [this, works, no, better, than, any, punch, yo...\n",
              "4    [purchased, this, thinking, maybe, need, speci...\n",
              "Name: reviewText, dtype: object"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 7
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "fOpTJYPRCogy"
      },
      "source": [
        "# Training the Word2Vec Model\n",
        "Train the model for reviews. Use a window of size 10 i.e. 10 words before the present word and 10 words ahead. A sentence with at least 2 words should only be considered, configure this using min_count parameter.\n",
        "\n",
        "Workers define how many CPU threads to be used."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "MjxwToPxCukY"
      },
      "source": [
        "\n",
        "**Initialize the model**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "a0VEbJYw7BT6"
      },
      "source": [
        "model = gensim.models.Word2Vec(\n",
        "    window=10,\n",
        "    min_count=2,\n",
        "    workers=5\n",
        "    \n",
        ")"
      ],
      "execution_count": 8,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "mf4cH_qmCyRa"
      },
      "source": [
        "**Build Vocabulary**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "mgVezRck8v_a"
      },
      "source": [
        "model.build_vocab(review_text,progress_per=100)"
      ],
      "execution_count": 9,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "WG0Y67l89Y3S",
        "outputId": "f674cf68-cf8d-441b-f8e7-c97e0342d9a2"
      },
      "source": [
        "model.corpus_count"
      ],
      "execution_count": 10,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "296337"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 10
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "CrUGNO98C3lh"
      },
      "source": [
        "**Train the Word2Vec Model**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "9H8JTSuv9cqC",
        "outputId": "7b09e20a-9d72-4709-950d-55c164cb231b"
      },
      "source": [
        "model.train(review_text, total_examples= model.corpus_count, epochs=model.epochs)"
      ],
      "execution_count": 11,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(91342671, 121496535)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 11
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "hdJYPlS9C8mw"
      },
      "source": [
        "**Saving and Loading model**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "hpMLrKH19hLx"
      },
      "source": [
        "# Saving model \n",
        "model.save(\"W2V-amazon-sports-outdoor5.model\")\n",
        "\n",
        "# Loading model\n",
        "#model = Word2Vec.load(\"W2V-amazon-sports-outdoor5.model.model\")"
      ],
      "execution_count": 12,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "pfsPQGBkDEdL"
      },
      "source": [
        "## Finding Similar Words and Similarity between words"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "JnWIxYSF-D3B",
        "outputId": "fd68e891-c017-48a9-9003-29f9cdb6e30a"
      },
      "source": [
        "model.wv.most_similar(\"bad\")"
      ],
      "execution_count": 13,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[('terrible', 0.6857774257659912),\n",
              " ('horrible', 0.6298424005508423),\n",
              " ('shabby', 0.6093147397041321),\n",
              " ('funny', 0.5464158058166504),\n",
              " ('upset', 0.5223544239997864),\n",
              " ('good', 0.5088357925415039),\n",
              " ('lame', 0.4958421587944031),\n",
              " ('greatest', 0.4954583942890167),\n",
              " ('poor', 0.4945375919342041),\n",
              " ('awful', 0.48850446939468384)]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 13
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "vDyXpJE3-apT",
        "outputId": "aa7b1478-6e71-4dc4-ddc9-9d555ff0e020"
      },
      "source": [
        "model.wv.most_similar(\"awful\")"
      ],
      "execution_count": 14,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[('terrible', 0.7020831108093262),\n",
              " ('horrible', 0.686090886592865),\n",
              " ('ugly', 0.6426234841346741),\n",
              " ('unpleasant', 0.639647364616394),\n",
              " ('overwhelming', 0.593713641166687),\n",
              " ('horrendous', 0.5878825187683105),\n",
              " ('overpowering', 0.5738529562950134),\n",
              " ('crummy', 0.5567977428436279),\n",
              " ('authentic', 0.5563438534736633),\n",
              " ('inferior', 0.539214015007019)]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 14
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "BvObEGiL-d1j",
        "outputId": "1a992d37-6c2f-42b0-c56e-3a995ad1b494"
      },
      "source": [
        "model.wv.similarity(w1='good',w2='great')"
      ],
      "execution_count": 15,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.79265124"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 15
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "LKqDnUT0-qdx",
        "outputId": "78c52a1e-7772-44f9-a0ae-06257fb77092"
      },
      "source": [
        "model.wv.similarity(w1='slow',w2='steady')"
      ],
      "execution_count": 16,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.38846418"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 16
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "pewE8ddiDS0I"
      },
      "source": [
        "# Further Reading\n",
        "You can read about gensim more at <br>https://radimrehurek.com/gensim/models/word2vec.html\n",
        "\n",
        "Explore other Datasets related to Amazon Reviews:<br> http://jmcauley.ucsd.edu/data/amazon/"
      ]
    }
  ]
}