HateXplain_BiRNN.ipynb

{
  "cells": [
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "qcpPVtw3YDKG"
      },
      "outputs": [],
      "source": [
        "import tensorflow as tf\n",
        "import numpy as np\n",
        "import pandas as pd\n",
        "import matplotlib.pyplot as plt\n",
        "import gensim\n",
        "import spacy\n",
        "import nltk\n",
        "\n",
        "from keras.models import Sequential\n",
        "from keras.layers import Embedding, Conv1D, MaxPooling1D, GRU, Dense, Dropout\n",
        "from nltk.tokenize import word_tokenize\n",
        "from wordcloud import WordCloud\n",
        "from gensim.models import Word2Vec\n",
        "from tensorflow.keras.utils import pad_sequences\n",
        "from tensorflow.keras.layers import Input, Embedding, LSTM, Dense, Dropout\n",
        "from tensorflow.keras.utils import pad_sequences\n",
        "from keras.preprocessing.text import Tokenizer\n",
        "from keras.layers import Embedding, Conv1D, MaxPooling1D, GRU, Dense, Dropout\n",
        "from keras.models import Sequential\n",
        "from keras.optimizers import RMSprop,Adam\n",
        "from keras.utils import to_categorical\n",
        "from sklearn.preprocessing import LabelEncoder\n",
        "from keras.preprocessing.text import Tokenizer\n",
        "     "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Nyut1gMlYfNv"
      },
      "source": []
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "1zyw0cXxYfkY",
        "outputId": "39038a62-5f0e-43e2-b700-fdb5882f6fde"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "--2023-04-13 05:52:00--  http://nlp.stanford.edu/data/glove.6B.zip\n",
            "Resolving nlp.stanford.edu (nlp.stanford.edu)... 171.64.67.140\n",
            "Connecting to nlp.stanford.edu (nlp.stanford.edu)|171.64.67.140|:80... connected.\n",
            "HTTP request sent, awaiting response... 302 Found\n",
            "Location: https://nlp.stanford.edu/data/glove.6B.zip [following]\n",
            "--2023-04-13 05:52:00--  https://nlp.stanford.edu/data/glove.6B.zip\n",
            "Connecting to nlp.stanford.edu (nlp.stanford.edu)|171.64.67.140|:443... connected.\n",
            "HTTP request sent, awaiting response... 301 Moved Permanently\n",
            "Location: https://downloads.cs.stanford.edu/nlp/data/glove.6B.zip [following]\n",
            "--2023-04-13 05:52:00--  https://downloads.cs.stanford.edu/nlp/data/glove.6B.zip\n",
            "Resolving downloads.cs.stanford.edu (downloads.cs.stanford.edu)... 171.64.64.22\n",
            "Connecting to downloads.cs.stanford.edu (downloads.cs.stanford.edu)|171.64.64.22|:443... connected.\n",
            "HTTP request sent, awaiting response... 200 OK\n",
            "Length: 862182613 (822M) [application/zip]\n",
            "Saving to: ‘glove.6B.zip’\n",
            "\n",
            "glove.6B.zip        100%[===================>] 822.24M  5.03MB/s    in 2m 38s  \n",
            "\n",
            "2023-04-13 05:54:39 (5.19 MB/s) - ‘glove.6B.zip’ saved [862182613/862182613]\n",
            "\n",
            "Archive:  glove.6B.zip\n",
            "  inflating: glove.6B.50d.txt        \n",
            "  inflating: glove.6B.100d.txt       \n",
            "  inflating: glove.6B.200d.txt       \n",
            "  inflating: glove.6B.300d.txt       \n"
          ]
        }
      ],
      "source": [
        "# Download the pre-trained GloVe word embeddings\n",
        "!wget http://nlp.stanford.edu/data/glove.6B.zip\n",
        "!unzip glove.6B.zip"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "Zv8orx2AV4sn",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "fcbe94ae-f11a-4c97-fba1-737b9b9c3c99"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Mounted at /content/drive\n"
          ]
        }
      ],
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/content/drive')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "NgP_627nXANz"
      },
      "outputs": [],
      "source": [
        "Base_path ='/content/drive/MyDrive/Data/hatexplain'"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "5rH8clD0X-0a"
      },
      "outputs": [],
      "source": [
        "df=pd.read_csv(Base_path+\"/HateXplain2.csv\")"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "Ews4spN0ZQzl",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 250
        },
        "outputId": "dbdb5856-e295-4e75-9f88-450b25c18ed8"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "   Unnamed: 0                      post_id  \\\n",
              "0           0  1179055004553900032_twitter   \n",
              "1           1  1179063826874032128_twitter   \n",
              "2           2  1178793830532956161_twitter   \n",
              "3           3  1179088797964763136_twitter   \n",
              "4           4  1179085312976445440_twitter   \n",
              "\n",
              "                                          rationales  \\\n",
              "0                                                 []   \n",
              "1                                                 []   \n",
              "2                                                 []   \n",
              "3  [[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,...   \n",
              "4  [[0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,...   \n",
              "\n",
              "                                         post_tokens Target_cat   Label_cat  \\\n",
              "0  ['i', 'dont', 'think', 'im', 'getting', 'my', ...       None      normal   \n",
              "1  ['we', 'cannot', 'continue', 'calling', 'ourse...       None      normal   \n",
              "2      ['nawt', 'yall', 'niggers', 'ignoring', 'me']    African      normal   \n",
              "3  ['<user>', 'i', 'am', 'bit', 'confused', 'coz'...      Asian  hatespeech   \n",
              "4  ['this', 'bitch', 'in', 'whataburger', 'eating...      Women  hatespeech   \n",
              "\n",
              "                                    final_rationales  \n",
              "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, ...  \n",
              "2                                    [0, 0, 0, 0, 0]  \n",
              "3  [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, ...  \n",
              "4  [0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, ...  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-0d8222d8-2d53-4025-9c2b-252bba30e6e0\">\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>Unnamed: 0</th>\n",
              "      <th>post_id</th>\n",
              "      <th>rationales</th>\n",
              "      <th>post_tokens</th>\n",
              "      <th>Target_cat</th>\n",
              "      <th>Label_cat</th>\n",
              "      <th>final_rationales</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>0</td>\n",
              "      <td>1179055004553900032_twitter</td>\n",
              "      <td>[]</td>\n",
              "      <td>['i', 'dont', 'think', 'im', 'getting', 'my', ...</td>\n",
              "      <td>None</td>\n",
              "      <td>normal</td>\n",
              "      <td>[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, ...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>1</td>\n",
              "      <td>1179063826874032128_twitter</td>\n",
              "      <td>[]</td>\n",
              "      <td>['we', 'cannot', 'continue', 'calling', 'ourse...</td>\n",
              "      <td>None</td>\n",
              "      <td>normal</td>\n",
              "      <td>[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, ...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>2</td>\n",
              "      <td>1178793830532956161_twitter</td>\n",
              "      <td>[]</td>\n",
              "      <td>['nawt', 'yall', 'niggers', 'ignoring', 'me']</td>\n",
              "      <td>African</td>\n",
              "      <td>normal</td>\n",
              "      <td>[0, 0, 0, 0, 0]</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>3</td>\n",
              "      <td>1179088797964763136_twitter</td>\n",
              "      <td>[[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,...</td>\n",
              "      <td>['&lt;user&gt;', 'i', 'am', 'bit', 'confused', 'coz'...</td>\n",
              "      <td>Asian</td>\n",
              "      <td>hatespeech</td>\n",
              "      <td>[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, ...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>4</td>\n",
              "      <td>1179085312976445440_twitter</td>\n",
              "      <td>[[0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,...</td>\n",
              "      <td>['this', 'bitch', 'in', 'whataburger', 'eating...</td>\n",
              "      <td>Women</td>\n",
              "      <td>hatespeech</td>\n",
              "      <td>[0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, ...</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-0d8222d8-2d53-4025-9c2b-252bba30e6e0')\"\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-0d8222d8-2d53-4025-9c2b-252bba30e6e0 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-0d8222d8-2d53-4025-9c2b-252bba30e6e0');\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": 6
        }
      ],
      "source": [
        "df.head()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "JsCAcU6qZSFv",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "3056b4d9-7275-4e7d-aa0a-de3c86eee815"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "<ipython-input-7-71b9c50e4ec9>:2: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  data['post_tokens'] = data['post_tokens'].apply(lambda x: eval(x))\n"
          ]
        }
      ],
      "source": [
        "data = df[['post_tokens',\t'Target_cat',\t'Label_cat',\t'final_rationales']]\n",
        "data['post_tokens'] = data['post_tokens'].apply(lambda x: eval(x))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "j0_OKfFSeDBX",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "ce995587-d3e6-492a-fcb8-c6bcd5e812cb"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0        [i, dont, think, im, getting, my, baby, them, ...\n",
              "1        [we, cannot, continue, calling, ourselves, fem...\n",
              "2                      [nawt, yall, niggers, ignoring, me]\n",
              "3        [<user>, i, am, bit, confused, coz, chinese, p...\n",
              "4        [this, bitch, in, whataburger, eating, a, burg...\n",
              "                               ...                        \n",
              "20143    [if, ur, still, on, twitter, tell, carlton, i,...\n",
              "20144    [when, i, first, got, on, here, and, said, i, ...\n",
              "20145    [was, macht, der, moslem, wenn, der, zion, geg...\n",
              "20146    [it, is, awful, look, at, world, demographics,...\n",
              "20147    [the, jewish, globalist, elite, have, only, im...\n",
              "Name: post_tokens, Length: 20148, dtype: object"
            ]
          },
          "metadata": {},
          "execution_count": 8
        }
      ],
      "source": [
        "data['post_tokens']"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "oSzfrd2iZWRq",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "845e02c1-0265-473b-ffc9-17e8ac2a4fa4"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "<ipython-input-9-751387b6647d>:5: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  data['post_tokens'].iloc[i] = sentence\n"
          ]
        }
      ],
      "source": [
        "for i in range(len(data)):\n",
        "  sentence =\"\"\n",
        "  for j in (data['post_tokens'].iloc[i]):\n",
        "    sentence += j +\" \"\n",
        "  data['post_tokens'].iloc[i] = sentence"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "7niaVL8sZbfN",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 206
        },
        "outputId": "f7e5e338-167e-43c8-8f23-56efd34da8db"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "                                         post_tokens Target_cat   Label_cat  \\\n",
              "0  i dont think im getting my baby them white 9 h...       None      normal   \n",
              "1  we cannot continue calling ourselves feminists...       None      normal   \n",
              "2                     nawt yall niggers ignoring me     African      normal   \n",
              "3  <user> i am bit confused coz chinese ppl can n...      Asian  hatespeech   \n",
              "4  this bitch in whataburger eating a burger with...      Women  hatespeech   \n",
              "\n",
              "                                    final_rationales  \n",
              "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, ...  \n",
              "2                                    [0, 0, 0, 0, 0]  \n",
              "3  [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, ...  \n",
              "4  [0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, ...  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-cefb73a8-8c64-4088-a514-d3cab012a659\">\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>post_tokens</th>\n",
              "      <th>Target_cat</th>\n",
              "      <th>Label_cat</th>\n",
              "      <th>final_rationales</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>i dont think im getting my baby them white 9 h...</td>\n",
              "      <td>None</td>\n",
              "      <td>normal</td>\n",
              "      <td>[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, ...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>we cannot continue calling ourselves feminists...</td>\n",
              "      <td>None</td>\n",
              "      <td>normal</td>\n",
              "      <td>[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, ...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>nawt yall niggers ignoring me</td>\n",
              "      <td>African</td>\n",
              "      <td>normal</td>\n",
              "      <td>[0, 0, 0, 0, 0]</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>&lt;user&gt; i am bit confused coz chinese ppl can n...</td>\n",
              "      <td>Asian</td>\n",
              "      <td>hatespeech</td>\n",
              "      <td>[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, ...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>this bitch in whataburger eating a burger with...</td>\n",
              "      <td>Women</td>\n",
              "      <td>hatespeech</td>\n",
              "      <td>[0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, ...</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-cefb73a8-8c64-4088-a514-d3cab012a659')\"\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-cefb73a8-8c64-4088-a514-d3cab012a659 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-cefb73a8-8c64-4088-a514-d3cab012a659');\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": 10
        }
      ],
      "source": [
        "data.head()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "czkOcJPlZed7"
      },
      "outputs": [],
      "source": [
        "label_map = {'hatespeech': 0, 'normal': 1, 'offensive': 2}\n",
        "labels = data[\"Label_cat\"].apply(lambda x: label_map[x])"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "fXKXbV3PZnWf",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "87d94546-ab1c-4cd6-a19a-222871e664f5"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0        1\n",
              "1        1\n",
              "2        1\n",
              "3        0\n",
              "4        0\n",
              "        ..\n",
              "20143    2\n",
              "20144    2\n",
              "20145    1\n",
              "20146    0\n",
              "20147    2\n",
              "Name: Label_cat, Length: 20148, dtype: int64"
            ]
          },
          "metadata": {},
          "execution_count": 12
        }
      ],
      "source": [
        "labels"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "FG2l6FrVZn8u"
      },
      "outputs": [],
      "source": [
        "max_words = 30000\n",
        "maxlen = 100"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "82mTy3bCZtAE"
      },
      "outputs": [],
      "source": [
        "# Tokenize the text data and convert it to sequences of integers\n",
        "tokenizer = Tokenizer(num_words=max_words)\n",
        "tokenizer.fit_on_texts(df['post_tokens'])\n",
        "sequences = tokenizer.texts_to_sequences(df['post_tokens'])\n",
        "X = pad_sequences(sequences, maxlen=maxlen)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "CJATdiRSZwj4"
      },
      "outputs": [],
      "source": [
        "# Define the target labels\n",
        "y = pd.get_dummies(df['Label_cat']).values"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "ePZrNaLMZx5s"
      },
      "outputs": [],
      "source": [
        "# Load the pre-trained GloVe embeddings\n",
        "embedding_dict = {}\n",
        "with open('glove.6B.100d.txt', encoding='utf-8') as f:\n",
        "    for line in f:\n",
        "        values = line.split()\n",
        "        word = values[0]\n",
        "        vector = np.asarray(values[1:], dtype='float32')\n",
        "        embedding_dict[word] = vector"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "R55g9LAvZ4UL"
      },
      "outputs": [],
      "source": [
        "# Initialize the embedding matrix with pre-trained GloVe embeddings\n",
        "word_index = tokenizer.word_index\n",
        "num_words = min(max_words, len(word_index) + 1)\n",
        "embedding_matrix = np.zeros((num_words, 100))\n",
        "for word, i in word_index.items():\n",
        "    if i >= max_words:\n",
        "        continue\n",
        "    embedding_vector = embedding_dict.get(word)\n",
        "    if embedding_vector is not None:\n",
        "        embedding_matrix[i] = embedding_vector"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "-pRip5F4aWq3"
      },
      "outputs": [],
      "source": [
        "# Define the input and output sizes\n",
        "input_size = 1000\n",
        "output_size = 10"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "GMe1QqdAZ850"
      },
      "outputs": [],
      "source": [
        "# Define the model architecture\n",
        "model = tf.keras.Sequential([\n",
        "    tf.keras.layers.Embedding(num_words, 100, weights=[embedding_matrix], input_length=maxlen, trainable=True),\n",
        "    tf.keras.layers.Bidirectional(tf.keras.layers.LSTM(64, return_sequences=True)),\n",
        "    tf.keras.layers.Bidirectional(tf.keras.layers.LSTM(32)),\n",
        "    tf.keras.layers.Dense(64, activation='relu'),\n",
        "    tf.keras.layers.Dense(3, activation='softmax')\n",
        "])"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "_GMNKQNSalyB"
      },
      "outputs": [],
      "source": [
        "# Compile the model\n",
        "model.compile(optimizer='adam',\n",
        "              loss='categorical_crossentropy',\n",
        "              metrics=['accuracy'])"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "KzRaC6fjbM6l",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "28b3aa5d-0bcf-4aaf-fde9-629cf29aae34"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"sequential\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " embedding (Embedding)       (None, 100, 100)          2885800   \n",
            "                                                                 \n",
            " bidirectional (Bidirectiona  (None, 100, 128)         84480     \n",
            " l)                                                              \n",
            "                                                                 \n",
            " bidirectional_1 (Bidirectio  (None, 64)               41216     \n",
            " nal)                                                            \n",
            "                                                                 \n",
            " dense (Dense)               (None, 64)                4160      \n",
            "                                                                 \n",
            " dense_1 (Dense)             (None, 3)                 195       \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 3,015,851\n",
            "Trainable params: 3,015,851\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ]
        }
      ],
      "source": [
        "# Print the model summary\n",
        "model.summary()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "75P9XrNgbO1A",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "63179f33-0364-4c49-f404-61847b96d6e0"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/10\n",
            "504/504 [==============================] - 171s 321ms/step - loss: 0.8670 - accuracy: 0.5975 - val_loss: 0.9885 - val_accuracy: 0.5035\n",
            "Epoch 2/10\n",
            "504/504 [==============================] - 159s 316ms/step - loss: 0.6030 - accuracy: 0.7551 - val_loss: 1.0541 - val_accuracy: 0.5112\n",
            "Epoch 3/10\n",
            "504/504 [==============================] - 164s 326ms/step - loss: 0.3991 - accuracy: 0.8499 - val_loss: 1.3122 - val_accuracy: 0.4950\n",
            "Epoch 4/10\n",
            "504/504 [==============================] - 162s 322ms/step - loss: 0.2622 - accuracy: 0.9064 - val_loss: 1.6331 - val_accuracy: 0.4876\n",
            "Epoch 5/10\n",
            "504/504 [==============================] - 163s 323ms/step - loss: 0.1699 - accuracy: 0.9409 - val_loss: 1.9946 - val_accuracy: 0.4794\n",
            "Epoch 6/10\n",
            "504/504 [==============================] - 162s 322ms/step - loss: 0.1238 - accuracy: 0.9595 - val_loss: 2.2023 - val_accuracy: 0.4777\n",
            "Epoch 7/10\n",
            "504/504 [==============================] - 161s 319ms/step - loss: 0.1009 - accuracy: 0.9648 - val_loss: 2.4173 - val_accuracy: 0.4834\n",
            "Epoch 8/10\n",
            "504/504 [==============================] - 160s 317ms/step - loss: 0.0740 - accuracy: 0.9736 - val_loss: 2.4327 - val_accuracy: 0.4824\n",
            "Epoch 9/10\n",
            "504/504 [==============================] - 162s 323ms/step - loss: 0.0574 - accuracy: 0.9784 - val_loss: 2.7505 - val_accuracy: 0.4782\n",
            "Epoch 10/10\n",
            "504/504 [==============================] - 163s 324ms/step - loss: 0.0531 - accuracy: 0.9808 - val_loss: 3.1076 - val_accuracy: 0.4801\n"
          ]
        }
      ],
      "source": [
        "history=model.fit(X, y, epochs=10, batch_size=32, validation_split=0.2)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "1cqDf80Oc6ow",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 472
        },
        "outputId": "eee3d162-c6f1-4324-dc45-bac5f3405992"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ],
      "source": [
        "# Plot the history of accuracy and loss\n",
        "plt.plot(history.history['accuracy'])\n",
        "plt.plot(history.history['val_accuracy'])\n",
        "plt.plot(history.history['loss'])\n",
        "plt.plot(history.history['val_loss'])\n",
        "plt.title('Model Performance')\n",
        "plt.xlabel('Epoch')\n",
        "plt.ylabel('Performance')\n",
        "plt.legend(['Train Accuracy', 'Validation Accuracy', 'Train Loss', 'Validation Loss'], loc='upper left')\n",
        "plt.show()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "Y8hcSNXZbUZl"
      },
      "outputs": [],
      "source": [
        "from sklearn.metrics import classification_report"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "bdf5kOizc_jy",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "89d4d030-b361-4f84-adb1-87f5a2cd1b66"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "630/630 [==============================] - 67s 101ms/step\n"
          ]
        }
      ],
      "source": [
        "y_pred = np.argmax(model.predict(X), axis=-1)\n",
        "\n",
        "# Convert the labels from one-hot encoding to integers\n",
        "y_test_int = np.argmax(y, axis=1)    "
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "_8D1FNvbdFii",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "a052a715-7015-4ef8-dafc-7deb0523b512"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.88      0.91      0.89      5934\n",
            "           1       0.90      0.93      0.92      7814\n",
            "           2       0.88      0.81      0.84      6400\n",
            "\n",
            "    accuracy                           0.89     20148\n",
            "   macro avg       0.89      0.89      0.89     20148\n",
            "weighted avg       0.89      0.89      0.89     20148\n",
            "\n"
          ]
        }
      ],
      "source": [
        "print(classification_report(y_test_int, y_pred))"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "model.save(\"/content/drive/MyDrive/Reproducibility/BiRNN.h5\")"
      ],
      "metadata": {
        "id": "COWkx6svTche"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "#calculate AUC of model\n",
        "auc = metrics.roc_auc_score(y_test, y_pred_proba)\n",
        "\n",
        "#print AUC score\n",
        "print(auc)\n",
        "\n",
        "0.5602104030579559"
      ],
      "metadata": {
        "id": "qK7IBt1UTxU0"
      },
      "execution_count": null,
      "outputs": []
    }
  ],
  "metadata": {
    "colab": {
      "provenance": []
    },
    "kernelspec": {
      "display_name": "Python 3",
      "name": "python3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "nbformat": 4,
  "nbformat_minor": 0
}