HindiFakeNews.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "HindiFakeNews.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "toc_visible": true,
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/sharikav/Fakenewsdetection/blob/master/HindiFakeNews.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "QeQPrU7P-jR9",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import pandas as pd\n",
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "2ISMd8aN-p5P",
        "colab_type": "code",
        "colab": {
          "resources": {
            "http://localhost:8080/nbextensions/google.colab/files.js": {
              "data": "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",
              "ok": true,
              "headers": [
                [
                  "content-type",
                  "application/javascript"
                ]
              ],
              "status": 200,
              "status_text": ""
            }
          },
          "base_uri": "https://localhost:8080/",
          "height": 105
        },
        "outputId": "0777b90b-1d2c-4347-86a7-f3f362857bed"
      },
      "source": [
        "from google.colab import files\n",
        "uploaded = files.upload()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "\n",
              "     <input type=\"file\" id=\"files-a93ad783-44dd-4983-b008-8648a506b321\" name=\"files[]\" multiple disabled\n",
              "        style=\"border:none\" />\n",
              "     <output id=\"result-a93ad783-44dd-4983-b008-8648a506b321\">\n",
              "      Upload widget is only available when the cell has been executed in the\n",
              "      current browser session. Please rerun this cell to enable.\n",
              "      </output>\n",
              "      <script src=\"/nbextensions/google.colab/files.js\"></script> "
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "Saving Cleaned_hindi_fake.csv to Cleaned_hindi_fake.csv\n",
            "Saving Cleaned_hindi_news.csv to Cleaned_hindi_news.csv\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "vjZEe7Nz_Qb1",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "true_file = \"Cleaned_hindi_news.csv\"\n",
        "fake_file = \"Cleaned_hindi_fake.csv\""
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "x6ltjYuD_qLK",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "df_true = pd.read_csv(true_file)\n",
        "df_fake = pd.read_csv(fake_file)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "D4iucQYA1mSx",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "P6xH8CEe1mzK",
        "colab_type": "text"
      },
      "source": [
        ""
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "9ScpdQ5r1nkK",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "cK4b9i4v_ywm",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 66
        },
        "outputId": "a3e20bc3-2a9a-4888-fe01-1f2e36be0dd7"
      },
      "source": [
        "df_true.isna().sum() #no nan rows"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Unnamed: 0    0\n",
              "Article       0\n",
              "dtype: int64"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 7
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "J8b4FG4y_-X3",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 66
        },
        "outputId": "761cce26-9888-409c-abdc-fc88b175855d"
      },
      "source": [
        "df_fake.isna().sum() #No nan rows"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Unnamed: 0    0\n",
              "0             0\n",
              "dtype: int64"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 8
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "hKOUqCDhAAtr",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 419
        },
        "outputId": "f5200a7b-2a2c-460c-b1b3-05d23ddabaa8"
      },
      "source": [
        "df_true"
      ],
      "execution_count": null,
      "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>Unnamed: 0</th>\n",
              "      <th>Article</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>0</td>\n",
              "      <td>सोनभद्र जिले में सोमवार को 20 और लोगों की जांच...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>1</td>\n",
              "      <td>Image Source : INSTAGRAM/TANUSRIDASGUPTA बालाज...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>2</td>\n",
              "      <td>एंटरटेनमेंट डेस्क, अमर उजाला, Updated Sun, 12 ...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>3</td>\n",
              "      <td>वाराणसी में बढ़ते कोरोना संक्रमित मरीजों की संख...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>4</td>\n",
              "      <td>Image Source : PTI जम्मू-कश्मीर भाजपा अध्यक्ष ...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>397</th>\n",
              "      <td>397</td>\n",
              "      <td>This Content has been blocked in your country ...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>398</th>\n",
              "      <td>398</td>\n",
              "      <td>एंटरटेनमेंट डेस्क, अमर उजाला, Updated Mon, 13 ...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>399</th>\n",
              "      <td>399</td>\n",
              "      <td>प्रतीकात्मक तस्वीर\\n\\nCovid-19 Live updates : ...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>400</th>\n",
              "      <td>400</td>\n",
              "      <td>Image Source : FILE Coronavirus cases in Harya...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>401</th>\n",
              "      <td>401</td>\n",
              "      <td>India Coronavirus Updates: भारत में कोरोना के ...</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>402 rows × 2 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "     Unnamed: 0                                            Article\n",
              "0             0  सोनभद्र जिले में सोमवार को 20 और लोगों की जांच...\n",
              "1             1  Image Source : INSTAGRAM/TANUSRIDASGUPTA बालाज...\n",
              "2             2  एंटरटेनमेंट डेस्क, अमर उजाला, Updated Sun, 12 ...\n",
              "3             3  वाराणसी में बढ़ते कोरोना संक्रमित मरीजों की संख...\n",
              "4             4  Image Source : PTI जम्मू-कश्मीर भाजपा अध्यक्ष ...\n",
              "..          ...                                                ...\n",
              "397         397  This Content has been blocked in your country ...\n",
              "398         398  एंटरटेनमेंट डेस्क, अमर उजाला, Updated Mon, 13 ...\n",
              "399         399  प्रतीकात्मक तस्वीर\\n\\nCovid-19 Live updates : ...\n",
              "400         400  Image Source : FILE Coronavirus cases in Harya...\n",
              "401         401  India Coronavirus Updates: भारत में कोरोना के ...\n",
              "\n",
              "[402 rows x 2 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 291
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "d2Ys8tPsAISK",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 406
        },
        "outputId": "669437e1-3a94-4b21-9d69-fbaf6d4d18c2"
      },
      "source": [
        "df_fake"
      ],
      "execution_count": null,
      "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>Unnamed: 0</th>\n",
              "      <th>0</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>0</td>\n",
              "      <td>कोरोनोवायरस फैलाने के लिए बर्तन चाटते मुसलमान ...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>1</td>\n",
              "      <td>इस्लामोफोबिक कथाओं के साथ वायरल हो रहे कोरोनोव...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>2</td>\n",
              "      <td>कराची में हिंदुओं ने राशन से इनकार किया, पाकिस...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>3</td>\n",
              "      <td>मुस्लिम व्यक्ति ने संगरोध से बचने के लिए पुलिस...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>4</td>\n",
              "      <td>Muslim cop beats Hindu priest in MP कहा जा रहा...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>436</th>\n",
              "      <td>436</td>\n",
              "      <td>चीन के विदेश मंत्रालय के प्रवक्ता ने ट्वीट किय...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>437</th>\n",
              "      <td>437</td>\n",
              "      <td>SARS महामारी के दौरान, टेक्सास के वैज्ञानिकों ...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>438</th>\n",
              "      <td>438</td>\n",
              "      <td>वायरस वास्तव में एक जैविक हथियार के रूप में ची...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>439</th>\n",
              "      <td>439</td>\n",
              "      <td>विरोध प्रदर्शन एक दौड़ युद्ध शुरू करने के लिए ...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>440</th>\n",
              "      <td>440</td>\n",
              "      <td>यहां तक ​​कि राष्ट्रपति ट्रम्प जैसे शीर्ष अधिक...</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>441 rows × 2 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "     Unnamed: 0                                                  0\n",
              "0             0  कोरोनोवायरस फैलाने के लिए बर्तन चाटते मुसलमान ...\n",
              "1             1  इस्लामोफोबिक कथाओं के साथ वायरल हो रहे कोरोनोव...\n",
              "2             2  कराची में हिंदुओं ने राशन से इनकार किया, पाकिस...\n",
              "3             3  मुस्लिम व्यक्ति ने संगरोध से बचने के लिए पुलिस...\n",
              "4             4  Muslim cop beats Hindu priest in MP कहा जा रहा...\n",
              "..          ...                                                ...\n",
              "436         436  चीन के विदेश मंत्रालय के प्रवक्ता ने ट्वीट किय...\n",
              "437         437  SARS महामारी के दौरान, टेक्सास के वैज्ञानिकों ...\n",
              "438         438  वायरस वास्तव में एक जैविक हथियार के रूप में ची...\n",
              "439         439  विरोध प्रदर्शन एक दौड़ युद्ध शुरू करने के लिए ...\n",
              "440         440  यहां तक ​​कि राष्ट्रपति ट्रम्प जैसे शीर्ष अधिक...\n",
              "\n",
              "[441 rows x 2 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 9
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Zz1WpFIDALR4",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "fake = df_fake[\"0\"]\n",
        "true = df_true[\"Article\"]"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ExMoAqq5AR_r",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 33
        },
        "outputId": "98c73e8d-6ae7-4bd4-dc53-fefa40d8b0ae"
      },
      "source": [
        "len(fake),len(true)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(441, 402)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 11
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "vG-HYjHvATsh",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 100
        },
        "outputId": "1460dfc5-e8dd-4c67-abe1-71c27ff7945c"
      },
      "source": [
        "print(\"Hindi Article\\n--------------------------\\n\")\n",
        "print(f\"Number of Fake Articles : {len(fake)} \")\n",
        "print(f\"Number of True Article : {len(true)}\")"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Hindi Article\n",
            "--------------------------\n",
            "\n",
            "Number of Fake Articles : 441 \n",
            "Number of True Article : 402\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "4xHlIxPNAlcf",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 197
        },
        "outputId": "2936fc98-1353-4050-a349-61f54a54e317"
      },
      "source": [
        "df_f = pd.DataFrame({'Article':fake})\n",
        "df_t = pd.DataFrame({'Article':true})\n",
        "df_t['category'] = 1\n",
        "df_f['category'] = 0\n",
        "df_t.to_csv(\"NewTrueh.csv\",index=False)\n",
        "df_f.to_csv(\"NewFakeh.csv\",index=False)\n",
        "df_f.head()"
      ],
      "execution_count": null,
      "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>Article</th>\n",
              "      <th>category</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>कोरोनोवायरस फैलाने के लिए बर्तन चाटते मुसलमान ...</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>इस्लामोफोबिक कथाओं के साथ वायरल हो रहे कोरोनोव...</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>कराची में हिंदुओं ने राशन से इनकार किया, पाकिस...</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>मुस्लिम व्यक्ति ने संगरोध से बचने के लिए पुलिस...</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>Muslim cop beats Hindu priest in MP कहा जा रहा...</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "                                             Article  category\n",
              "0  कोरोनोवायरस फैलाने के लिए बर्तन चाटते मुसलमान ...         0\n",
              "1  इस्लामोफोबिक कथाओं के साथ वायरल हो रहे कोरोनोव...         0\n",
              "2  कराची में हिंदुओं ने राशन से इनकार किया, पाकिस...         0\n",
              "3  मुस्लिम व्यक्ति ने संगरोध से बचने के लिए पुलिस...         0\n",
              "4  Muslim cop beats Hindu priest in MP कहा जा रहा...         0"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 13
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "adM7-Vo8Bx-1",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 406
        },
        "outputId": "3fdb1574-9b7f-4a73-c25a-69cb53cdc928"
      },
      "source": [
        "new_articles = list(df_t['Article'])\n",
        "category1 = list(df_t['category'])\n",
        "category2 = list(df_f['category'])\n",
        "new_articles.extend(list(df_f['Article']))\n",
        "df = pd.DataFrame({'Article':new_articles,'Category':category1 + category2})\n",
        "df"
      ],
      "execution_count": null,
      "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>Article</th>\n",
              "      <th>Category</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>सोनभद्र जिले में सोमवार को 20 और लोगों की जांच...</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>Image Source : INSTAGRAM/TANUSRIDASGUPTA बालाज...</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>एंटरटेनमेंट डेस्क, अमर उजाला, Updated Sun, 12 ...</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>वाराणसी में बढ़ते कोरोना संक्रमित मरीजों की संख...</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>Image Source : PTI जम्मू-कश्मीर भाजपा अध्यक्ष ...</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>838</th>\n",
              "      <td>चीन के विदेश मंत्रालय के प्रवक्ता ने ट्वीट किय...</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>839</th>\n",
              "      <td>SARS महामारी के दौरान, टेक्सास के वैज्ञानिकों ...</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>840</th>\n",
              "      <td>वायरस वास्तव में एक जैविक हथियार के रूप में ची...</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>841</th>\n",
              "      <td>विरोध प्रदर्शन एक दौड़ युद्ध शुरू करने के लिए ...</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>842</th>\n",
              "      <td>यहां तक ​​कि राष्ट्रपति ट्रम्प जैसे शीर्ष अधिक...</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>843 rows × 2 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "                                               Article  Category\n",
              "0    सोनभद्र जिले में सोमवार को 20 और लोगों की जांच...         1\n",
              "1    Image Source : INSTAGRAM/TANUSRIDASGUPTA बालाज...         1\n",
              "2    एंटरटेनमेंट डेस्क, अमर उजाला, Updated Sun, 12 ...         1\n",
              "3    वाराणसी में बढ़ते कोरोना संक्रमित मरीजों की संख...         1\n",
              "4    Image Source : PTI जम्मू-कश्मीर भाजपा अध्यक्ष ...         1\n",
              "..                                                 ...       ...\n",
              "838  चीन के विदेश मंत्रालय के प्रवक्ता ने ट्वीट किय...         0\n",
              "839  SARS महामारी के दौरान, टेक्सास के वैज्ञानिकों ...         0\n",
              "840  वायरस वास्तव में एक जैविक हथियार के रूप में ची...         0\n",
              "841  विरोध प्रदर्शन एक दौड़ युद्ध शुरू करने के लिए ...         0\n",
              "842  यहां तक ​​कि राष्ट्रपति ट्रम्प जैसे शीर्ष अधिक...         0\n",
              "\n",
              "[843 rows x 2 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 14
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "8aWdzSlbA9ob",
        "colab_type": "text"
      },
      "source": [
        "# Data Cleaning\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "5WOhwxVSBFZT",
        "colab_type": "text"
      },
      "source": [
        "##Removes the following things for further cleaning of the articles:\n",
        "\n",
        "\n",
        "*   PUNCTUATIONS\n",
        "*   EMOJI'S\n",
        "*   DATES\n",
        "*   LINKS\n",
        "*   EMAILS\n",
        "*   URLS\n",
        "*   NUMBERS\n",
        "\n",
        "\n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "qB8cjAyEBVpR",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "  df = df.replace(r'[.\\w]{3,}@[.\\w]{5,}','', regex=True) #email\n",
        "  df = df.replace(r'https?://[.\\w]{3,}','', regex=True) #urls\n",
        "  df = df.replace(r'www.[.\\w]{3,}','', regex=True) #urls\n",
        "  df = df.replace(r'[0-9]','', regex=True) #numbers\n",
        "  df = df.replace(r'\\xa0',' ', regex=True) #escape sequences\n",
        "  df = df.replace(r'\\n',' ', regex=True)\n",
        "  df = df.replace(r'\\t','', regex=True)\n",
        "  df = df.replace(r'\\r','', regex=True)\n",
        "  df = df.replace(r'\\s\\s\\s','', regex=True)\n",
        "  #df = df.replace(r'\\u','',regex=True)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "IcAgl2GEBoed",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "df.to_csv(\"cleaned.csv\") "
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "1QHgxi8qfpfK",
        "colab_type": "text"
      },
      "source": [
        "## **DATA PREPROCESSING**"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "4ZrdmGsMfy-K",
        "colab_type": "text"
      },
      "source": [
        "*   Removing punctuation\n",
        "*   Tokenization\n",
        "*   Lemmatizing\n",
        "*   Stopwords\n",
        "*   Made a new csv file with final data"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "6vm9im2jeRC3",
        "colab_type": "text"
      },
      "source": [
        "We used stemming for hindi"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "DpMGyhuoClRC",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 50
        },
        "outputId": "0ac1b7e0-cb86-46b1-f0ac-b866c8bfe03a"
      },
      "source": [
        "import nltk \n",
        "nltk.download('punkt')\n",
        "\n",
        "punctuations = '''!()-[]{};:'\"\\,<>./?@#$%^&*_~'''\n",
        "\n",
        "\n",
        "def remove_punct(my_str):\n",
        "  no_punct = \"\"\n",
        "  for char in my_str:\n",
        "    if char not in punctuations:\n",
        "        no_punct = no_punct + char\n",
        "  return no_punct\n",
        "\n",
        "def generate_stem_words(word):\n",
        "\t\tsuffixes = {\n",
        "    1: [u\"ो\",u\"े\",u\"ू\",u\"ु\",u\"ी\",u\"ि\",u\"ा\"],\n",
        "    2: [u\"कर\",u\"ाओ\",u\"िए\",u\"ाई\",u\"ाए\",u\"ने\",u\"नी\",u\"ना\",u\"ते\",u\"ीं\",u\"ती\",u\"ता\",u\"ाँ\",u\"ां\",u\"ों\",u\"ें\"],\n",
        "    3: [u\"ाकर\",u\"ाइए\",u\"ाईं\",u\"ाया\",u\"ेगी\",u\"ेगा\",u\"ोगी\",u\"ोगे\",u\"ाने\",u\"ाना\",u\"ाते\",u\"ाती\",u\"ाता\",u\"तीं\",u\"ाओं\",u\"ाएं\",u\"ुओं\",u\"ुएं\",u\"ुआं\"],\n",
        "    4: [u\"ाएगी\",u\"ाएगा\",u\"ाओगी\",u\"ाओगे\",u\"एंगी\",u\"ेंगी\",u\"एंगे\",u\"ेंगे\",u\"ूंगी\",u\"ूंगा\",u\"ातीं\",u\"नाओं\",u\"नाएं\",u\"ताओं\",u\"ताएं\",u\"ियाँ\",u\"ियों\",u\"ियां\"],\n",
        "    5: [u\"ाएंगी\",u\"ाएंगे\",u\"ाऊंगी\",u\"ाऊंगा\",u\"ाइयाँ\",u\"ाइयों\",u\"ाइयां\"],\n",
        "}\n",
        "  \n",
        "\t\tfor L in 5, 4, 3, 2, 1:\n",
        "\t\t\tif len(word) > L + 1:\n",
        "\t\t\t\tfor suf in suffixes[L]:\n",
        "\t\t\t\t\t#print type(suf),type(word),word,suf\n",
        "\t\t\t\t\tif word.endswith(suf):\n",
        "\t\t\t\t\t\t#print 'h'\n",
        "\t\t\t\t\t\treturn word[:-L]\n",
        "\t\treturn word\n",
        "def generate(text):\n",
        "  new=[]\n",
        "  for w in text:\n",
        "    word = generate_stem_words(w)\n",
        "    new.append(word)\n",
        "  return \" \".join(new)\n",
        "with open(\"stopwords.txt\") as f:\n",
        "  all_stopwords = f.read().split(\"\\n\")\n",
        "\n",
        "def remove_stop_words(text_list):\n",
        "  new = [text for text in text_list if not (text in all_stopwords)]\n",
        "  return \" \".join(new)\n",
        "hi = remove_stop_words(nltk.word_tokenize(df['Article'][0]))\n",
        "\n",
        "def preprocess(df):\n",
        "  df['punctuated'] = df['Article'].apply(remove_punct)\n",
        "  df['stemmed'] = df['punctuated'].apply(lambda x:generate(nltk.word_tokenize(x)))\n",
        "  df['stopwords'] = df['stemmed'].apply(lambda x:remove_stop_words(nltk.word_tokenize(x)))\n",
        "  return df\n",
        "  "
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[nltk_data] Downloading package punkt to /root/nltk_data...\n",
            "[nltk_data]   Unzipping tokenizers/punkt.zip.\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ezEXpBeiC3UV",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "\n",
        "df = preprocess(df)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "0wmyADbWDoiu",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 590
        },
        "outputId": "75383b91-9561-4200-d822-faa52a248d51"
      },
      "source": [
        "df"
      ],
      "execution_count": null,
      "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>Article</th>\n",
              "      <th>Category</th>\n",
              "      <th>punctuated</th>\n",
              "      <th>stemmed</th>\n",
              "      <th>stopwords</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>सोनभद्र जिले में सोमवार को  और लोगों की जांच र...</td>\n",
              "      <td>1</td>\n",
              "      <td>सोनभद्र जिले में सोमवार को  और लोगों की जांच र...</td>\n",
              "      <td>सोनभद्र जिल में सोमवार को और लोग की जांच रिपोर...</td>\n",
              "      <td>सोनभद्र जिल सोमवार लोग जांच रिपोर्ट कोरो पाजिट...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>Image Source : INSTAGRAM/TANUSRIDASGUPTA बालाज...</td>\n",
              "      <td>1</td>\n",
              "      <td>Image Source  INSTAGRAMTANUSRIDASGUPTA बालाजी ...</td>\n",
              "      <td>Image Source INSTAGRAMTANUSRIDASGUPTA बालाज टे...</td>\n",
              "      <td>Image Source INSTAGRAMTANUSRIDASGUPTA बालाज टे...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>एंटरटेनमेंट डेस्क, अमर उजाला, Updated Sun,  Ju...</td>\n",
              "      <td>1</td>\n",
              "      <td>एंटरटेनमेंट डेस्क अमर उजाला Updated Sun  Jul  ...</td>\n",
              "      <td>एंटरटेनमेंट डेस्क अमर उजाल Updated Sun Jul PM ...</td>\n",
              "      <td>एंटरटेनमेंट डेस्क अमर उजाल Updated Sun Jul PM ...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>वाराणसी में बढ़ते कोरोना संक्रमित मरीजों की संख...</td>\n",
              "      <td>1</td>\n",
              "      <td>वाराणसी में बढ़ते कोरोना संक्रमित मरीजों की संख...</td>\n",
              "      <td>वाराणस में बढ़ कोरो संक्रमित मरीज की संख्य को द...</td>\n",
              "      <td>वाराणस बढ़ कोरो संक्रमित मरीज संख्य देख जिल प्र...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>Image Source : PTI जम्मू-कश्मीर भाजपा अध्यक्ष ...</td>\n",
              "      <td>1</td>\n",
              "      <td>Image Source  PTI जम्मूकश्मीर भाजपा अध्यक्ष रव...</td>\n",
              "      <td>Image Source PTI जम्मूकश्मीर भाजप अध्यक्ष रविं...</td>\n",
              "      <td>Image Source PTI जम्मूकश्मीर भाजप अध्यक्ष रविं...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>838</th>\n",
              "      <td>चीन के विदेश मंत्रालय के प्रवक्ता ने ट्वीट किय...</td>\n",
              "      <td>0</td>\n",
              "      <td>चीन के विदेश मंत्रालय के प्रवक्ता ने ट्वीट किय...</td>\n",
              "      <td>चीन के विदेश मंत्रालय के प्रवक् ने ट्वीट किय क...</td>\n",
              "      <td>चीन विदेश मंत्रालय प्रवक् ट्वीट किय संभव अमेरि...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>839</th>\n",
              "      <td>SARS महामारी के दौरान, टेक्सास के वैज्ञानिकों ...</td>\n",
              "      <td>0</td>\n",
              "      <td>SARS महामारी के दौरान टेक्सास के वैज्ञानिकों क...</td>\n",
              "      <td>SARS महामार के दौरान टेक्सास के वैज्ञानिक की ए...</td>\n",
              "      <td>SARS महामार दौरान टेक्सास वैज्ञानिक टीम वायरस ...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>840</th>\n",
              "      <td>वायरस वास्तव में एक जैविक हथियार के रूप में ची...</td>\n",
              "      <td>0</td>\n",
              "      <td>वायरस वास्तव में एक जैविक हथियार के रूप में ची...</td>\n",
              "      <td>वायरस वास्तव में एक जैविक हथियार के रूप में ची...</td>\n",
              "      <td>वायरस वास्तव जैविक हथियार रूप ची वैज्ञानिक द्व...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>841</th>\n",
              "      <td>विरोध प्रदर्शन एक दौड़ युद्ध शुरू करने के लिए ...</td>\n",
              "      <td>0</td>\n",
              "      <td>विरोध प्रदर्शन एक दौड़ युद्ध शुरू करने के लिए ...</td>\n",
              "      <td>विरोध प्रदर्शन एक दौड़ युद्ध शुर कर के लिए किय...</td>\n",
              "      <td>विरोध प्रदर्शन दौड़ युद्ध शुर किय गय ताक आग लॉ...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>842</th>\n",
              "      <td>यहां तक ​​कि राष्ट्रपति ट्रम्प जैसे शीर्ष अधिक...</td>\n",
              "      <td>0</td>\n",
              "      <td>यहां तक ​​कि राष्ट्रपति ट्रम्प जैसे शीर्ष अधिक...</td>\n",
              "      <td>यह तक ​​क राष्ट्रपत ट्रम्प जैस शीर्ष अधिकार से...</td>\n",
              "      <td>​​क राष्ट्रपत ट्रम्प जैस शीर्ष अधिकार जानकार म...</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>843 rows × 5 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "                                               Article  ...                                          stopwords\n",
              "0    सोनभद्र जिले में सोमवार को  और लोगों की जांच र...  ...  सोनभद्र जिल सोमवार लोग जांच रिपोर्ट कोरो पाजिट...\n",
              "1    Image Source : INSTAGRAM/TANUSRIDASGUPTA बालाज...  ...  Image Source INSTAGRAMTANUSRIDASGUPTA बालाज टे...\n",
              "2    एंटरटेनमेंट डेस्क, अमर उजाला, Updated Sun,  Ju...  ...  एंटरटेनमेंट डेस्क अमर उजाल Updated Sun Jul PM ...\n",
              "3    वाराणसी में बढ़ते कोरोना संक्रमित मरीजों की संख...  ...  वाराणस बढ़ कोरो संक्रमित मरीज संख्य देख जिल प्र...\n",
              "4    Image Source : PTI जम्मू-कश्मीर भाजपा अध्यक्ष ...  ...  Image Source PTI जम्मूकश्मीर भाजप अध्यक्ष रविं...\n",
              "..                                                 ...  ...                                                ...\n",
              "838  चीन के विदेश मंत्रालय के प्रवक्ता ने ट्वीट किय...  ...  चीन विदेश मंत्रालय प्रवक् ट्वीट किय संभव अमेरि...\n",
              "839  SARS महामारी के दौरान, टेक्सास के वैज्ञानिकों ...  ...  SARS महामार दौरान टेक्सास वैज्ञानिक टीम वायरस ...\n",
              "840  वायरस वास्तव में एक जैविक हथियार के रूप में ची...  ...  वायरस वास्तव जैविक हथियार रूप ची वैज्ञानिक द्व...\n",
              "841  विरोध प्रदर्शन एक दौड़ युद्ध शुरू करने के लिए ...  ...  विरोध प्रदर्शन दौड़ युद्ध शुर किय गय ताक आग लॉ...\n",
              "842  यहां तक ​​कि राष्ट्रपति ट्रम्प जैसे शीर्ष अधिक...  ...  ​​क राष्ट्रपत ट्रम्प जैस शीर्ष अधिकार जानकार म...\n",
              "\n",
              "[843 rows x 5 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 19
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "8mkzdYCNDqCQ",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 154
        },
        "outputId": "7374125c-9bf7-4fc9-a443-18fc0c9676b0"
      },
      "source": [
        "df['punctuated'][5]"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "application/vnd.google.colaboratory.intrinsic": {
              "type": "string"
            },
            "text/plain": [
              "'कुपोषण को लेकर UN की रिपोर्ट पर बोले पूर्व खाद्य सचिव सरकार के  किलो अनाज देने से नहीं चलेगा काम पीठ ने कहा कि एक बार इसका निदान हो जाए तो यह सभी अस्पतालों को प्रेषित किया जाना चाहिए अदालत ने इसका समाधान तलाशने के लिए आईसीएमआर को  जुलाई तक का वक्त दिया है और हलफनामे के जरिए अदालत को सूचित करने को कहा है मानव व्यवहार और संबद्ध विज्ञान संस्थान आईएचबीएस ने वकील तुषार सन्नू के जरिए अदालत को सूचित किया था कि उसके यहां भर्ती मानसिक रूप से बीमार लोगों की कोरोना वायरस की जांच करने में दो बाधाएं हैं इसके बाद अदालत का आदेश आया सन्नू ने कहा कि  जून के आईसीएमआर के परामर्श के अनुसार जिस व्यक्ति की कोविड की जांच होनी है उसे सरकार द्वारा जारी पहचान प्रमाण देना होगा और उसके पास वैध फोन नंबर होना चाहिए ताकि उसका और उसके संपर्क में आए लोगों का पता लगाया जा सके उन्होंने कहा कि अधिकतर मानसिक रूप से बीमार व्यक्ति दोनों में से कोई भी शर्त पूरी नहीं करते हैं लिहाजा उनकी कोरोना वायरस की जांच नहीं की जा सकती है आईएचबीएएस ने सुझाव दिया कि आईसीएमआर के दिशानिर्देशों में मानसिक रूप से बीमार बेघर लोगों की परेशानी के मद्देनजर उचित संशोधन किये जाएं  उच्च न्यायालय वकील गौरव कुमार बंसल की याचिका पर सुनवाई कर रहा था  उन्होंने अपनी याचिका में राष्ट्रीय राजधानी में मानसिक रूप से बीमार बेघर लोगों की कोविड जांच के लिए आईसीएमआर और दिल्ली सरकार को दिशानिर्देश जारी करने के लिए निर्देश देने का अनुरोध किया है दिल्ली सरकार ने अपने हलफनामे में कहा है कि वह आईसीएमआर के  जून के परामर्श का पालन कर रही है  VIDEOप्लाज़्मा लेने वाले ज़्यादा देने वाले कम अब दिल्ली प्लाज़्मा बैंक ने बनाई ये नई व्यवस्था'"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 303
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ymfJb-faD3nn",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 154
        },
        "outputId": "73d2d934-9abe-40ff-e4d2-3d4e9508b153"
      },
      "source": [
        "df['stemmed'][5]"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "application/vnd.google.colaboratory.intrinsic": {
              "type": "string"
            },
            "text/plain": [
              "'कुपोषण को ले UN की रिपोर्ट पर बोल पूर्व खाद्य सचिव सरकार के किल अनाज दे से नह चल काम पीठ ने कह कि एक बार इसक निदान हो जाए तो यह सभ अस्पताल को प्रेषित किय जा चाह अदालत ने इसक समाधान तलाश के लिए आईसीएमआर को जुल तक का वक्त दिय है और हलफनाम के जर अदालत को सूचित कर को कह है मानव व्यवहार और संबद्ध विज्ञान संस्थान आईएचबीएस ने वकील तुषार सन्न के जर अदालत को सूचित किय था कि उसक यह भर् मानसिक रूप से बीमार लोग की कोरो वायरस की जांच कर में दो बाध हैं इसक बाद अदालत का आदेश आय सन्न ने कह कि जून के आईसीएमआर के परामर्श के अनुसार जिस व्यक्त की कोविड की जांच हो है उस सरकार द्वार जार पहचान प्रमाण दे होग और उसक पास वैध फोन नंबर हो चाह ताक उसक और उसक संपर्क में आए लोग का पत लग जा सक उन्हों कह कि अधिकतर मानसिक रूप से बीमार व्यक्त दोन में से कोई भी शर्त पूर नह कर हैं लिहाज उनक कोरो वायरस की जांच नह की जा सक है आईएचबीएएस ने सुझाव दिय कि आईसीएमआर के दिशानिर्देश में मानसिक रूप से बीमार बेघर लोग की परेशा के मद्देनजर उचित संशोधन किय जाएं उच्च न्यायालय वकील गौरव कुमार बंसल की याचिक पर सुनव कर रह था उन्हों अप याचिक में राष्ट्रीय राजधा में मानसिक रूप से बीमार बेघर लोग की कोविड जांच के लिए आईसीएमआर और दिल्ल सरकार को दिशानिर्देश जार कर के लिए निर्देश दे का अनुरोध किय है दिल्ल सरकार ने अप हलफनाम में कह है कि वह आईसीएमआर के जून के परामर्श का पालन कर रह है VIDEOप्लाज़्म ले वाल ज़्याद दे वाल कम अब दिल्ल प्लाज़्म बैंक ने बन ये नई व्यवस्थ'"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 304
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "FuHjhqHXD7C_",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 154
        },
        "outputId": "00d13a4e-ba47-40e4-a9a1-f59502c07c88"
      },
      "source": [
        "df['stopwords'][5]"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "application/vnd.google.colaboratory.intrinsic": {
              "type": "string"
            },
            "text/plain": [
              "'कुपोषण ले UN रिपोर्ट बोल पूर्व खाद्य सचिव सरकार किल अनाज दे नह चल काम पीठ कह बार इसक निदान जाए सभ अस्पताल प्रेषित किय चाह अदालत इसक समाधान तलाश आईसीएमआर जुल वक्त दिय हलफनाम जर अदालत सूचित कह मानव व्यवहार संबद्ध विज्ञान संस्थान आईएचबीएस वकील तुषार सन्न जर अदालत सूचित किय उसक भर् मानसिक रूप बीमार लोग कोरो वायरस जांच बाध इसक अदालत आदेश आय सन्न कह जून आईसीएमआर परामर्श अनुसार व्यक्त कोविड जांच सरकार द्वार जार पहचान प्रमाण दे होग उसक पास वैध फोन नंबर चाह ताक उसक उसक संपर्क आए लोग पत लग सक कह अधिकतर मानसिक रूप बीमार व्यक्त दोन शर्त पूर नह लिहाज उनक कोरो वायरस जांच नह सक आईएचबीएएस सुझाव दिय आईसीएमआर दिशानिर्देश मानसिक रूप बीमार बेघर लोग परेशा मद्देनजर उचित संशोधन किय जाएं उच्च न्यायालय वकील गौरव कुमार बंसल याचिक सुनव रह अप याचिक राष्ट्रीय राजधा मानसिक रूप बीमार बेघर लोग कोविड जांच आईसीएमआर दिल्ल सरकार दिशानिर्देश जार निर्देश दे अनुरोध किय दिल्ल सरकार अप हलफनाम कह आईसीएमआर जून परामर्श पालन रह VIDEOप्लाज़्म ले वाल ज़्याद दे वाल कम अब दिल्ल प्लाज़्म बैंक बन नई व्यवस्थ'"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 305
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "GlAm8PJaED00",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "df.to_csv(\"cleaned_hindi.csv\")"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "wLdUsX8Gf_MU",
        "colab_type": "text"
      },
      "source": [
        "## **VECTORISATION**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "zvt0eno8EQoI",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from sklearn.feature_extraction.text import TfidfVectorizer\n",
        "vectorizer = TfidfVectorizer(\n",
        "                            \n",
        "                            analyzer='word',\n",
        "                            ngram_range=(1, 3),\n",
        "                            max_features=2000\n",
        "                            #max_df=0.5 # Verwendet im ML-Kurs unter Preprocessing                   \n",
        "                            )\n",
        "\n",
        "X = vectorizer.fit_transform(df['stopwords']).toarray() # remember to use the original X_train set\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "GUrTurAQEcPT",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 136
        },
        "outputId": "e7de693e-17aa-47a7-fe71-ffc8f841e2dd"
      },
      "source": [
        "vectorizer.get_feature_names"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<bound method CountVectorizer.get_feature_names of TfidfVectorizer(analyzer='word', binary=False, decode_error='strict',\n",
              "                dtype=<class 'numpy.float64'>, encoding='utf-8',\n",
              "                input='content', lowercase=True, max_df=1.0, max_features=2000,\n",
              "                min_df=1, ngram_range=(1, 3), norm='l2', preprocessor=None,\n",
              "                smooth_idf=True, stop_words=None, strip_accents=None,\n",
              "                sublinear_tf=False, token_pattern='(?u)\\\\b\\\\w\\\\w+\\\\b',\n",
              "                tokenizer=None, use_idf=True, vocabulary=None)>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 308
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "h1sor-67EgSJ",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "2236c54b-a284-4f82-bec4-fea831d02210"
      },
      "source": [
        "vectorizer.vocabulary_"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "{'मव': 1422,\n",
              " 'आई': 405,\n",
              " 'इनम': 502,\n",
              " 'रम': 1604,\n",
              " 'पर': 1177,\n",
              " 'सर': 1926,\n",
              " 'आठ': 441,\n",
              " 'कर': 710,\n",
              " 'इनक': 501,\n",
              " 'अल': 377,\n",
              " 'बर': 1290,\n",
              " 'सग': 1872,\n",
              " 'हर': 1983,\n",
              " 'मल': 1386,\n",
              " 'रह': 1680,\n",
              " 'इसक': 515,\n",
              " 'एमओ': 628,\n",
              " 'एसक': 640,\n",
              " 'उप': 568,\n",
              " 'अब': 337,\n",
              " 'सबस': 1898,\n",
              " 'रस': 1678,\n",
              " 'इत': 494,\n",
              " 'मर': 1356,\n",
              " 'गई': 772,\n",
              " 'वस': 1805,\n",
              " 'गए': 799,\n",
              " 'जबक': 933,\n",
              " 'मध': 1346,\n",
              " 'यर': 1475,\n",
              " 'टर': 963,\n",
              " 'इल': 504,\n",
              " 'चल': 897,\n",
              " 'रम रम': 1633,\n",
              " 'रम पर': 1620,\n",
              " 'पर सर': 1199,\n",
              " 'मल रह': 1418,\n",
              " 'रह इसक': 1689,\n",
              " 'अब सबस': 349,\n",
              " 'इसक रम': 524,\n",
              " 'रम गई': 1616,\n",
              " 'गए जबक': 805,\n",
              " 'यर टर': 1477,\n",
              " 'इल चल': 507,\n",
              " 'चल रह': 901,\n",
              " 'इल चल रह': 508,\n",
              " 'image': 98,\n",
              " 'source': 198,\n",
              " 'एट': 618,\n",
              " 'समय': 1920,\n",
              " 'यरस': 1482,\n",
              " 'दस': 1065,\n",
              " 'तक': 994,\n",
              " 'मह': 1426,\n",
              " 'यक': 1450,\n",
              " 'अम': 361,\n",
              " 'बच': 1237,\n",
              " 'चन': 882,\n",
              " 'अभ': 352,\n",
              " 'ऐश': 646,\n",
              " 'वर': 1789,\n",
              " 'आर': 472,\n",
              " 'अन': 299,\n",
              " 'पम': 1174,\n",
              " 'एक': 596,\n",
              " 'समथ': 1919,\n",
              " 'गय': 820,\n",
              " 'कप': 698,\n",
              " 'तन': 1008,\n",
              " 'चप': 891,\n",
              " 'बत': 1268,\n",
              " 'पहल': 1220,\n",
              " 'ऑक': 654,\n",
              " 'जन': 924,\n",
              " 'वल': 1800,\n",
              " 'लग': 1737,\n",
              " 'उन': 549,\n",
              " 'बई': 1231,\n",
              " 'वन': 1785,\n",
              " 'टल': 976,\n",
              " 'एडम': 620,\n",
              " 'उनक': 558,\n",
              " 'लक': 1732,\n",
              " 'षण': 1840,\n",
              " 'कस': 742,\n",
              " 'दग': 1044,\n",
              " 'नए': 1076,\n",
              " 'एप': 623,\n",
              " 'नह': 1131,\n",
              " 'कह': 743,\n",
              " 'कम': 703,\n",
              " 'शन': 1816,\n",
              " 'अस': 387,\n",
              " 'पत': 1154,\n",
              " 'लड': 1752,\n",
              " 'सल': 1946,\n",
              " 'गर': 864,\n",
              " 'सक': 1854,\n",
              " 'इसस': 527,\n",
              " 'अप': 322,\n",
              " 'जर': 946,\n",
              " 'रत': 1552,\n",
              " 'आग': 434,\n",
              " 'गहर': 875,\n",
              " 'भर': 1315,\n",
              " 'पड': 1151,\n",
              " 'मच': 1330,\n",
              " 'मदद': 1344,\n",
              " 'उसक': 586,\n",
              " 'अगर': 269,\n",
              " 'सह': 1952,\n",
              " 'डर': 987,\n",
              " 'हम': 1974,\n",
              " 'सभ': 1907,\n",
              " 'जल': 951,\n",
              " 'पस': 1210,\n",
              " 'रतलब': 1572,\n",
              " 'यल': 1530,\n",
              " 'आन': 450,\n",
              " 'तस': 1029,\n",
              " 'करक': 724,\n",
              " 'हल': 1994,\n",
              " 'छल': 905,\n",
              " 'आय': 470,\n",
              " 'एमस': 630,\n",
              " 'इडल': 492,\n",
              " 'इन': 495,\n",
              " 'सप': 1890,\n",
              " 'टफ': 961,\n",
              " 'रख': 1541,\n",
              " 'एग': 608,\n",
              " 'full': 82,\n",
              " 'coverage': 55,\n",
              " 'image source': 99,\n",
              " 'यरस रह': 1522,\n",
              " 'रह अब': 1684,\n",
              " 'मह यक': 1440,\n",
              " 'अम बच': 363,\n",
              " 'बच चन': 1238,\n",
              " 'चन अभ': 883,\n",
              " 'ऐश वर': 647,\n",
              " 'वर आर': 1791,\n",
              " 'रम गए': 1618,\n",
              " 'अन पम': 306,\n",
              " 'गई एक': 782,\n",
              " 'एक टर': 599,\n",
              " 'गय अब': 827,\n",
              " 'ऑक जन': 655,\n",
              " 'जन वल': 926,\n",
              " 'टल एडम': 977,\n",
              " 'गय उनक': 835,\n",
              " 'सबस पहल': 1903,\n",
              " 'लक षण': 1733,\n",
              " 'शन अस': 1817,\n",
              " 'अस पत': 388,\n",
              " 'पत गय': 1161,\n",
              " 'गय टर': 843,\n",
              " 'इसस पहल': 528,\n",
              " 'पहल अप': 1221,\n",
              " 'जर रत': 948,\n",
              " 'रत नह': 1556,\n",
              " 'नह रह': 1141,\n",
              " 'आग कह': 435,\n",
              " 'कर मच': 717,\n",
              " 'रह अभ': 1685,\n",
              " 'नह हम': 1145,\n",
              " 'हम सभ': 1980,\n",
              " 'रह जल': 1705,\n",
              " 'यर करक': 1476,\n",
              " 'हल लक': 1995,\n",
              " 'षण सक': 1847,\n",
              " 'गय सभ': 858,\n",
              " 'पर आय': 1181,\n",
              " 'इडल इन': 493,\n",
              " 'full coverage': 83,\n",
              " 'अम बच चन': 364,\n",
              " 'बच चन अभ': 1239,\n",
              " 'ऐश वर आर': 648,\n",
              " 'ऑक जन वल': 656,\n",
              " 'अस पत गय': 393,\n",
              " 'जर रत नह': 949,\n",
              " 'हल लक षण': 1996,\n",
              " 'टरट': 972,\n",
              " 'नम': 1107,\n",
              " 'अमर': 367,\n",
              " 'उज': 535,\n",
              " 'updated': 244,\n",
              " 'sun': 208,\n",
              " 'jul': 133,\n",
              " 'pm': 178,\n",
              " 'शल': 1830,\n",
              " 'टरट नम': 973,\n",
              " 'नम अमर': 1108,\n",
              " 'अमर उज': 368,\n",
              " 'उज updated': 536,\n",
              " 'updated sun': 247,\n",
              " 'sun jul': 209,\n",
              " 'jul pm': 138,\n",
              " 'वर बच': 1794,\n",
              " 'टरट नम अमर': 974,\n",
              " 'नम अमर उज': 1109,\n",
              " 'अमर उज updated': 369,\n",
              " 'updated sun jul': 248,\n",
              " 'sun jul pm': 210,\n",
              " 'वर बच चन': 1795,\n",
              " 'णस': 993,\n",
              " 'बढ़': 1306,\n",
              " 'रश': 1676,\n",
              " 'सन': 1888,\n",
              " 'सख': 1871,\n",
              " 'बरत': 1293,\n",
              " 'लय': 1764,\n",
              " 'ठक': 980,\n",
              " 'एम': 624,\n",
              " 'शर': 1829,\n",
              " 'लन': 1755,\n",
              " 'अध': 280,\n",
              " 'कड': 677,\n",
              " 'मज': 1333,\n",
              " 'नल': 1121,\n",
              " 'नगर': 1094,\n",
              " 'गम': 819,\n",
              " 'रव': 1660,\n",
              " 'खर': 771,\n",
              " 'बज': 1244,\n",
              " 'बह': 1298,\n",
              " 'सड': 1880,\n",
              " 'वजह': 1779,\n",
              " 'उड': 539,\n",
              " 'लर': 1769,\n",
              " 'सम': 1910,\n",
              " 'तथ': 1007,\n",
              " 'उल': 583,\n",
              " 'घन': 877,\n",
              " 'सच': 1874,\n",
              " 'नजर': 1099,\n",
              " 'बस': 1296,\n",
              " 'लव': 1770,\n",
              " 'सरक': 1933,\n",
              " 'इव': 512,\n",
              " 'यद': 1462,\n",
              " 'यम': 1467,\n",
              " 'गल': 868,\n",
              " 'सफ': 1892,\n",
              " 'उठ': 538,\n",
              " 'जम': 939,\n",
              " 'आद': 444,\n",
              " 'आव': 476,\n",
              " 'मक': 1324,\n",
              " 'जगह': 919,\n",
              " 'छत': 903,\n",
              " 'कब': 700,\n",
              " 'उनस': 567,\n",
              " 'नर': 1120,\n",
              " 'कल': 727,\n",
              " 'कन': 697,\n",
              " 'पहच': 1217,\n",
              " 'रमण': 1638,\n",
              " 'एड': 619,\n",
              " 'एसप': 643,\n",
              " 'एचय': 612,\n",
              " 'रच': 1545,\n",
              " 'पहर': 1219,\n",
              " 'एस': 637,\n",
              " 'अपन': 332,\n",
              " 'दव': 1064,\n",
              " 'करव': 726,\n",
              " 'रम मर': 1625,\n",
              " 'रश सन': 1677,\n",
              " 'कड उन': 681,\n",
              " 'नगर गम': 1095,\n",
              " 'अप अप': 323,\n",
              " 'अप कह': 325,\n",
              " 'बज बज': 1246,\n",
              " 'बह बज': 1303,\n",
              " 'एम कह': 625,\n",
              " 'उल घन': 584,\n",
              " 'यम उल': 1470,\n",
              " 'कर कर': 711,\n",
              " 'कल सर': 730,\n",
              " 'रह सभ': 1726,\n",
              " 'पर रह': 1193,\n",
              " 'अप रमण': 329,\n",
              " 'कड अन': 678,\n",
              " 'लन कर': 1756,\n",
              " 'pti': 185,\n",
              " 'कश': 739,\n",
              " 'जप': 931,\n",
              " 'बड': 1248,\n",
              " 'खबर': 769,\n",
              " 'धव': 1072,\n",
              " 'आत': 443,\n",
              " 'उत': 540,\n",
              " 'तर': 1018,\n",
              " 'गठन': 812,\n",
              " 'अश': 385,\n",
              " 'कव': 731,\n",
              " 'वज': 1777,\n",
              " 'बढ': 1251,\n",
              " 'तव': 1028,\n",
              " 'गलव': 872,\n",
              " 'आध': 449,\n",
              " 'वय': 1788,\n",
              " 'दर': 1054,\n",
              " 'आज': 439,\n",
              " 'रभ': 1603,\n",
              " 'बन': 1282,\n",
              " 'नज': 1098,\n",
              " 'बड़': 1305,\n",
              " 'रक': 1537,\n",
              " 'तहस': 1032,\n",
              " 'हट': 1963,\n",
              " 'वहन': 1808,\n",
              " 'हन': 1971,\n",
              " 'मत': 1337,\n",
              " 'अगल': 271,\n",
              " 'वजन': 1778,\n",
              " 'आवश': 478,\n",
              " 'source pti': 202,\n",
              " 'जम कश': 940,\n",
              " 'अध यक': 290,\n",
              " 'रव रम': 1669,\n",
              " 'रव यरस': 1668,\n",
              " 'यरस रम': 1511,\n",
              " 'अन अब': 301,\n",
              " 'इन सभ': 500,\n",
              " 'रह बत': 1712,\n",
              " 'बत रम': 1276,\n",
              " 'रम गय': 1619,\n",
              " 'उत तर': 541,\n",
              " 'इल गय': 506,\n",
              " 'गय छल': 840,\n",
              " 'छल मह': 911,\n",
              " 'मह सच': 1444,\n",
              " 'अन कह': 302,\n",
              " 'मल बढ': 1409,\n",
              " 'रह कल': 1699,\n",
              " 'कह उनक': 747,\n",
              " 'रम यरस': 1632,\n",
              " 'चल गय': 899,\n",
              " 'गय हम': 862,\n",
              " 'मल लग': 1419,\n",
              " 'लग बढ': 1744,\n",
              " 'बढ रह': 1264,\n",
              " 'अब यरस': 344,\n",
              " 'यरस मल': 1503,\n",
              " 'लग गई': 1741,\n",
              " 'आद कह': 445,\n",
              " 'कह कड': 748,\n",
              " 'पर वहन': 1194,\n",
              " 'कह गय': 751,\n",
              " 'अन मत': 309,\n",
              " 'आद रह': 447,\n",
              " 'आवश यक': 479,\n",
              " 'रख गय': 1542,\n",
              " 'image source pti': 102,\n",
              " 'यरस रम गए': 1513,\n",
              " 'मल लग बढ': 1420,\n",
              " 'लग बढ रह': 1745,\n",
              " 'कह कड उन': 749,\n",
              " 'un': 240,\n",
              " 'अद': 277,\n",
              " 'लत': 1753,\n",
              " 'तल': 1027,\n",
              " 'आईस': 410,\n",
              " 'एमआर': 627,\n",
              " 'वक': 1774,\n",
              " 'नव': 1123,\n",
              " 'यवह': 1533,\n",
              " 'बद': 1279,\n",
              " 'नस': 1129,\n",
              " 'आए': 412,\n",
              " 'कतर': 693,\n",
              " 'घर': 879,\n",
              " 'मद': 1342,\n",
              " 'उच': 534,\n",
              " 'धन': 1069,\n",
              " 'जध': 923,\n",
              " 'नई': 1075,\n",
              " 'यवस': 1532,\n",
              " 'षण un': 1841,\n",
              " 'un सच': 241,\n",
              " 'सच सरक': 1877,\n",
              " 'सरक अन': 1934,\n",
              " 'अन नह': 304,\n",
              " 'नह चल': 1138,\n",
              " 'कह इसक': 746,\n",
              " 'सभ अस': 1908,\n",
              " 'अद लत': 278,\n",
              " 'आईस एमआर': 411,\n",
              " 'पर मर': 1188,\n",
              " 'उसक पर': 587,\n",
              " 'पर आए': 1179,\n",
              " 'पत लग': 1170,\n",
              " 'लग सक': 1747,\n",
              " 'सक कह': 1860,\n",
              " 'अध कतर': 281,\n",
              " 'नह उनक': 1134,\n",
              " 'यरस नह': 1498,\n",
              " 'नह सक': 1144,\n",
              " 'मद नजर': 1343,\n",
              " 'रह अप': 1683,\n",
              " 'षण un सच': 1842,\n",
              " 'un सच सरक': 242,\n",
              " 'सच सरक अन': 1878,\n",
              " 'सरक अन नह': 1935,\n",
              " 'अन नह चल': 305,\n",
              " 'सभ अस पत': 1909,\n",
              " 'रद': 1573,\n",
              " 'गण': 815,\n",
              " 'सव': 1949,\n",
              " 'पक': 1147,\n",
              " 'वव': 1802,\n",
              " 'हज': 1957,\n",
              " 'चव': 902,\n",
              " 'उसस': 589,\n",
              " 'नय': 1115,\n",
              " 'उपच': 570,\n",
              " 'सद': 1884,\n",
              " 'अबतक': 351,\n",
              " 'गत': 816,\n",
              " 'खतरन': 767,\n",
              " 'पश': 1208,\n",
              " 'तम': 1012,\n",
              " 'जस': 954,\n",
              " 'टक': 958,\n",
              " 'रल': 1659,\n",
              " 'ओड': 665,\n",
              " 'गढ': 814,\n",
              " 'असम': 398,\n",
              " 'सगढ': 1873,\n",
              " 'लद': 1754,\n",
              " 'एकएक': 604,\n",
              " 'इट': 489,\n",
              " 'नग': 1093,\n",
              " 'अर': 371,\n",
              " 'डम': 986,\n",
              " 'हव': 1997,\n",
              " 'षद': 1848,\n",
              " 'यव': 1531,\n",
              " 'सत': 1881,\n",
              " 'video': 254,\n",
              " 'रमण मल': 1654,\n",
              " 'मल आए': 1388,\n",
              " 'वव लय': 1803,\n",
              " 'लय अन': 1765,\n",
              " 'अन रत': 315,\n",
              " 'बढ हज': 1267,\n",
              " 'रत शन': 1569,\n",
              " 'शन यरस': 1822,\n",
              " 'यरस मह': 1505,\n",
              " 'मह सबस': 1445,\n",
              " 'सबस रभ': 1906,\n",
              " 'बन गय': 1284,\n",
              " 'गय अम': 829,\n",
              " 'हज अध': 1958,\n",
              " 'नय मल': 1116,\n",
              " 'लय बत': 1767,\n",
              " 'मर उपच': 1364,\n",
              " 'रह जबक': 1703,\n",
              " 'वर अध': 1790,\n",
              " 'अध कह': 284,\n",
              " 'कह अब': 745,\n",
              " 'सद मर': 1885,\n",
              " 'मर वस': 1384,\n",
              " 'खतरन यरस': 768,\n",
              " 'यरस अब': 1486,\n",
              " 'मह इसक': 1429,\n",
              " 'मध रद': 1349,\n",
              " 'तर रद': 1021,\n",
              " 'तम लन': 1014,\n",
              " 'कर टक': 713,\n",
              " 'चल रद': 900,\n",
              " 'छत सगढ': 904,\n",
              " 'बस इट': 1297,\n",
              " 'शन बह': 1821,\n",
              " 'गए कड': 803,\n",
              " 'सबस मल': 1905,\n",
              " 'मल मह': 1414,\n",
              " 'लन जर': 1758,\n",
              " 'रद रम': 1582,\n",
              " 'रम बढ': 1623,\n",
              " 'बढ गई': 1255,\n",
              " 'गई इसक': 780,\n",
              " 'रद मर': 1579,\n",
              " 'मर हर': 1385,\n",
              " 'मल रम': 1417,\n",
              " 'अन पर': 307,\n",
              " 'पर षद': 1197,\n",
              " 'रत यरस': 1559,\n",
              " 'रमण मल बढ': 1655,\n",
              " 'नय मल आए': 1117,\n",
              " 'उत तर रद': 542,\n",
              " 'रम बढ गई': 1624,\n",
              " 'ओर': 666,\n",
              " 'रण': 1548,\n",
              " 'पह': 1211,\n",
              " 'दक': 1043,\n",
              " 'अरव': 375,\n",
              " 'ऑनल': 659,\n",
              " 'कट': 675,\n",
              " 'लगभग': 1748,\n",
              " 'पय': 1175,\n",
              " 'सस': 1950,\n",
              " 'नन': 1104,\n",
              " 'तरर': 1025,\n",
              " 'बय': 1288,\n",
              " 'इसम': 525,\n",
              " 'लखनऊ': 1736,\n",
              " 'अहमद': 401,\n",
              " 'जयप': 945,\n",
              " 'यरस रमण': 1515,\n",
              " 'रमण सर': 1657,\n",
              " 'नए मर': 1078,\n",
              " 'मर आए': 1362,\n",
              " 'आए रम': 424,\n",
              " 'गई रम': 792,\n",
              " 'हज गई': 1959,\n",
              " 'मर अस': 1360,\n",
              " 'पत गई': 1160,\n",
              " 'जबक अब': 934,\n",
              " 'अब हज': 350,\n",
              " 'नए मल': 1080,\n",
              " 'आए जबक': 418,\n",
              " 'कड गय': 685,\n",
              " 'गय यरस': 850,\n",
              " 'यरस कड': 1489,\n",
              " 'उन रत': 554,\n",
              " 'कम कम': 707,\n",
              " 'धन यव': 1070,\n",
              " 'ऑनल इन': 660,\n",
              " 'बत मह': 1274,\n",
              " 'बय कह': 1289,\n",
              " 'बढ तर': 1259,\n",
              " 'नए मर आए': 1079,\n",
              " 'आए रम गई': 425,\n",
              " 'मर अस पत': 1361,\n",
              " 'अस पत गई': 392,\n",
              " 'नए मल आए': 1081,\n",
              " 'यरस कड उन': 1490,\n",
              " 'mon': 155,\n",
              " 'am': 4,\n",
              " 'updated mon': 245,\n",
              " 'mon jul': 156,\n",
              " 'jul am': 134,\n",
              " 'updated mon jul': 246,\n",
              " 'ऐस': 650,\n",
              " 'छह': 916,\n",
              " 'हस': 1998,\n",
              " 'षज': 1839,\n",
              " 'बदल': 1280,\n",
              " 'घट': 876,\n",
              " 'मतलब': 1340,\n",
              " 'आश': 480,\n",
              " 'जत': 921,\n",
              " 'आख': 432,\n",
              " 'सपर': 1891,\n",
              " 'डल': 989,\n",
              " 'बत गय': 1271,\n",
              " 'गय मर': 846,\n",
              " 'सक सक': 1869,\n",
              " 'हस पत': 1999,\n",
              " 'बत अब': 1269,\n",
              " 'लग रह': 1746,\n",
              " 'एग इसक': 609,\n",
              " 'रमण दर': 1644,\n",
              " 'बत रह': 1278,\n",
              " 'रह कड': 1694,\n",
              " 'गई अब': 779,\n",
              " 'अब कर': 338,\n",
              " 'दर गई': 1056,\n",
              " 'इसक मतलब': 522,\n",
              " 'हर रम': 1990,\n",
              " 'मर रह': 1383,\n",
              " 'अध सक': 298,\n",
              " 'मह पहल': 1438,\n",
              " 'मह कर': 1430,\n",
              " 'अब मह': 343,\n",
              " 'एक सपर': 603,\n",
              " 'बत मर': 1273,\n",
              " 'रह कड उन': 1695,\n",
              " 'file': 78,\n",
              " 'photo': 177,\n",
              " 'कदम': 695,\n",
              " 'चक': 880,\n",
              " 'भव': 1320,\n",
              " 'हफ': 1973,\n",
              " 'अच': 273,\n",
              " 'शभर': 1827,\n",
              " 'शत': 1812,\n",
              " 'source file': 199,\n",
              " 'file photo': 79,\n",
              " 'कर रह': 721,\n",
              " 'रह हर': 1731,\n",
              " 'गलव बह': 873,\n",
              " 'बह कड': 1301,\n",
              " 'बत कड': 1270,\n",
              " 'मल गए': 1403,\n",
              " 'यक यरस': 1454,\n",
              " 'सरक सव': 1945,\n",
              " 'रमण बढ': 1651,\n",
              " 'सर नए': 1927,\n",
              " 'मल आन': 1394,\n",
              " 'आन रम': 454,\n",
              " 'गई तक': 786,\n",
              " 'तक बढ': 997,\n",
              " 'गई लय': 794,\n",
              " 'एक मल': 601,\n",
              " 'बढ गए': 1257,\n",
              " 'मल मल': 1413,\n",
              " 'मल रत': 1415,\n",
              " 'रत शत': 1566,\n",
              " 'image source file': 100,\n",
              " 'यरस रमण मल': 1519,\n",
              " 'यरस मल बढ': 1504,\n",
              " 'सर नए मल': 1928,\n",
              " 'नए मल आन': 1082,\n",
              " 'मल आन रम': 1396,\n",
              " 'आन रम गई': 455,\n",
              " 'तक बढ गई': 998,\n",
              " 'मल बढ गए': 1410,\n",
              " 'coronavirus': 44,\n",
              " 'updates': 249,\n",
              " 'यतन': 1461,\n",
              " 'उनम': 566,\n",
              " 'covid': 56,\n",
              " 'in': 103,\n",
              " 'hindi': 93,\n",
              " 'ist': 120,\n",
              " 'एज': 613,\n",
              " 'नक': 1085,\n",
              " 'आगर': 438,\n",
              " 'एन': 621,\n",
              " 'उब': 579,\n",
              " 'तमब': 1015,\n",
              " 'अवध': 382,\n",
              " 'जनपद': 928,\n",
              " 'नसभ': 1130,\n",
              " 'टन': 960,\n",
              " 'दद': 1046,\n",
              " 'लज': 1751,\n",
              " 'new': 160,\n",
              " 'positive': 181,\n",
              " 'cases': 28,\n",
              " 'and': 5,\n",
              " 'deaths': 67,\n",
              " 'have': 90,\n",
              " 'been': 18,\n",
              " 'reported': 189,\n",
              " 'till': 227,\n",
              " 'today': 230,\n",
              " 'total': 233,\n",
              " 'number': 163,\n",
              " 'of': 167,\n",
              " 'at': 16,\n",
              " 'including': 114,\n",
              " 'active': 0,\n",
              " 'state': 205,\n",
              " 'health': 91,\n",
              " 'department': 69,\n",
              " 'ani': 7,\n",
              " 'july': 140,\n",
              " 'कवर': 732,\n",
              " 'जव': 953,\n",
              " 'रत मक': 1557,\n",
              " 'मक तस': 1325,\n",
              " 'तस coronavirus': 1030,\n",
              " 'coronavirus updates': 51,\n",
              " 'updates रत': 252,\n",
              " 'आन रमण': 456,\n",
              " 'गए मर': 807,\n",
              " 'मर गई': 1367,\n",
              " 'लय रव': 1768,\n",
              " 'रव बह': 1667,\n",
              " 'बह आठ': 1299,\n",
              " 'आठ बज': 442,\n",
              " 'अन छल': 303,\n",
              " 'छल नए': 907,\n",
              " 'जबक मर': 936,\n",
              " 'गय अभ': 828,\n",
              " 'रह अध': 1681,\n",
              " 'अभ मर': 360,\n",
              " 'मर दर': 1373,\n",
              " 'दर रत': 1059,\n",
              " 'छल गई': 906,\n",
              " 'सबस अध': 1899,\n",
              " 'मह तम': 1432,\n",
              " 'एकएक यक': 605,\n",
              " 'coronavirus covid': 47,\n",
              " 'covid updates': 60,\n",
              " 'updates in': 250,\n",
              " 'in hindi': 104,\n",
              " 'hindi jul': 94,\n",
              " 'jul ist': 135,\n",
              " 'ist मह': 130,\n",
              " 'मह यरस': 1441,\n",
              " 'यरस नय': 1496,\n",
              " 'मल एज': 1399,\n",
              " 'एज अन': 614,\n",
              " 'अन मह': 313,\n",
              " 'रमण नए': 1645,\n",
              " 'मल आय': 1398,\n",
              " 'हज गय': 1960,\n",
              " 'गय रद': 852,\n",
              " 'इसक नक': 521,\n",
              " 'नक jul': 1086,\n",
              " 'आए एज': 415,\n",
              " 'यरस नए': 1494,\n",
              " 'बढ गय': 1258,\n",
              " 'मर इल': 1363,\n",
              " 'इसक मल': 523,\n",
              " 'गए नक': 806,\n",
              " 'रद यरस': 1581,\n",
              " 'आए रद': 422,\n",
              " 'मल गई': 1402,\n",
              " 'गई छल': 783,\n",
              " 'गई jul': 773,\n",
              " 'रद नए': 1577,\n",
              " 'अन रद': 317,\n",
              " 'छल मल': 910,\n",
              " 'आए सक': 426,\n",
              " 'रव रमण': 1670,\n",
              " 'बढ़ गई': 1307,\n",
              " 'गई सम': 797,\n",
              " 'वर मह': 1796,\n",
              " 'मह नए': 1435,\n",
              " 'मह गई': 1431,\n",
              " 'रद रमण': 1586,\n",
              " 'सक रम': 1867,\n",
              " 'तक गई': 996,\n",
              " 'आग रह': 437,\n",
              " 'रह गय': 1701,\n",
              " 'तमब नगर': 1016,\n",
              " 'छल मर': 909,\n",
              " 'यरस वजह': 1525,\n",
              " 'तम नगर': 1013,\n",
              " 'अध नक': 287,\n",
              " 'नक बत': 1092,\n",
              " 'बत छल': 1272,\n",
              " 'अप यरस': 327,\n",
              " 'हर कल': 1986,\n",
              " 'रह उनक': 1692,\n",
              " 'ist जस': 128,\n",
              " 'new covid': 161,\n",
              " 'covid positive': 58,\n",
              " 'positive cases': 182,\n",
              " 'cases and': 29,\n",
              " 'and deaths': 6,\n",
              " 'reported in': 190,\n",
              " 'total number': 234,\n",
              " 'number of': 164,\n",
              " 'of cases': 168,\n",
              " 'active cases': 1,\n",
              " 'state health': 206,\n",
              " 'health department': 92,\n",
              " 'ani ani': 8,\n",
              " 'ani july': 11,\n",
              " 'ist छल': 127,\n",
              " 'सबस नए': 1901,\n",
              " 'मल अब': 1387,\n",
              " 'अब मल': 342,\n",
              " 'गए इसक': 801,\n",
              " 'रत मक तस': 1558,\n",
              " 'मक तस coronavirus': 1326,\n",
              " 'मल आन रमण': 1397,\n",
              " 'आन रमण मल': 457,\n",
              " 'बह आठ बज': 1300,\n",
              " 'कड अन छल': 679,\n",
              " 'छल नए मल': 908,\n",
              " 'दर रत शत': 1060,\n",
              " 'मह तम लन': 1433,\n",
              " 'covid updates in': 61,\n",
              " 'updates in hindi': 251,\n",
              " 'in hindi jul': 105,\n",
              " 'hindi jul ist': 95,\n",
              " 'jul ist मह': 137,\n",
              " 'यरस नय मल': 1497,\n",
              " 'यरस रमण नए': 1516,\n",
              " 'रमण नए मल': 1646,\n",
              " 'नक jul ist': 1087,\n",
              " 'मल आए एज': 1390,\n",
              " 'आए एज अन': 416,\n",
              " 'यरस नए मल': 1495,\n",
              " 'मल आए रम': 1392,\n",
              " 'मल आए रद': 1391,\n",
              " 'गई jul ist': 774,\n",
              " 'रद नए मल': 1578,\n",
              " 'मल आए सक': 1393,\n",
              " 'मह नए मल': 1436,\n",
              " 'आए सक रम': 427,\n",
              " 'jul ist जस': 136,\n",
              " 'covid positive cases': 59,\n",
              " 'positive cases and': 183,\n",
              " 'cases and deaths': 30,\n",
              " 'total number of': 235,\n",
              " 'number of cases': 165,\n",
              " 'state health department': 207,\n",
              " 'ani ani july': 9,\n",
              " 'सबस नए मल': 1902,\n",
              " 'ऑर': 662,\n",
              " 'गन': 817,\n",
              " 'इज': 485,\n",
              " 'अलगअलग': 380,\n",
              " 'आपक': 460,\n",
              " 'समस': 1924,\n",
              " 'आम': 468,\n",
              " 'डब': 983,\n",
              " 'एचओ': 611,\n",
              " 'आस': 482,\n",
              " 'उतर': 545,\n",
              " 'थक': 1033,\n",
              " 'महस': 1449,\n",
              " 'रज': 1546,\n",
              " 'हमल': 1982,\n",
              " 'रदर': 1590,\n",
              " 'ऊपर': 593,\n",
              " 'इसल': 526,\n",
              " 'वच': 1776,\n",
              " 'यत': 1460,\n",
              " 'इटल': 490,\n",
              " 'कठ': 676,\n",
              " 'तकल': 1002,\n",
              " 'दब': 1051,\n",
              " 'समझ': 1918,\n",
              " 'सब': 1895,\n",
              " 'आमत': 469,\n",
              " 'आसप': 483,\n",
              " 'पल': 1202,\n",
              " 'इस': 514,\n",
              " 'पकड': 1148,\n",
              " 'षण यरस': 1845,\n",
              " 'रत रह': 1564,\n",
              " 'इज शन': 487,\n",
              " 'रह मल': 1714,\n",
              " 'यरस लक': 1523,\n",
              " 'रह ऐस': 1693,\n",
              " 'रमण बच': 1650,\n",
              " 'बच सक': 1243,\n",
              " 'सक यरस': 1866,\n",
              " 'आपक बत': 462,\n",
              " 'बत यरस': 1275,\n",
              " 'डब एचओ': 984,\n",
              " 'अन रम': 318,\n",
              " 'पर कर': 1182,\n",
              " 'रम अध': 1607,\n",
              " 'रम यक': 1631,\n",
              " 'बन रह': 1286,\n",
              " 'समस रह': 1925,\n",
              " 'रह यरस': 1716,\n",
              " 'बदल रह': 1281,\n",
              " 'अगर आपक': 270,\n",
              " 'रह टर': 1707,\n",
              " 'रह जर': 1704,\n",
              " 'सह यक': 1953,\n",
              " 'सक इसल': 1858,\n",
              " 'दर गय': 1058,\n",
              " 'पल ऑक': 1203,\n",
              " 'ऑक टर': 657,\n",
              " 'आख रक': 433,\n",
              " 'तर यरस': 1020,\n",
              " 'यरस बढ': 1501,\n",
              " 'कम इन': 705,\n",
              " 'लक षण यरस': 1735,\n",
              " 'मल बढ रह': 1411,\n",
              " 'यरस लक षण': 1524,\n",
              " 'पल ऑक टर': 1204,\n",
              " 'सट': 1879,\n",
              " 'णय': 992,\n",
              " 'अहम': 400,\n",
              " 'तहत': 1031,\n",
              " 'एसड': 641,\n",
              " 'सदर': 1886,\n",
              " 'अख': 267,\n",
              " 'एशन': 636,\n",
              " 'सहमत': 1954,\n",
              " 'एसड एम': 642,\n",
              " 'एस एशन': 638,\n",
              " 'रव नह': 1665,\n",
              " 'कर लग': 722,\n",
              " 'लन नह': 1759,\n",
              " 'नह एग': 1135,\n",
              " 'कहर': 762,\n",
              " 'जह': 957,\n",
              " 'ist यरस': 131,\n",
              " 'रत लग': 1565,\n",
              " 'अब रम': 347,\n",
              " 'सक मल': 1864,\n",
              " 'representational': 191,\n",
              " 'उपय': 575,\n",
              " 'एसएफ': 639,\n",
              " 'शहर': 1837,\n",
              " 'बल': 1295,\n",
              " 'थकव': 1034,\n",
              " 'षक': 1838,\n",
              " 'इक': 484,\n",
              " 'रवक': 1674,\n",
              " 'representational image': 192,\n",
              " 'बढ मल': 1263,\n",
              " 'दर गए': 1057,\n",
              " 'लग कड': 1739,\n",
              " 'आए नम': 419,\n",
              " 'अप कर': 324,\n",
              " 'पत कह': 1159,\n",
              " 'रह अन': 1682,\n",
              " 'लगभग रत': 1749,\n",
              " 'लग कड उन': 1740,\n",
              " 'लगभग रत शत': 1750,\n",
              " 'मजद': 1334,\n",
              " 'नत': 1101,\n",
              " 'इम': 503,\n",
              " 'खड': 764,\n",
              " 'पद': 1171,\n",
              " 'भय': 1314,\n",
              " 'मग': 1329,\n",
              " 'दल': 1063,\n",
              " 'वश': 1804,\n",
              " 'जड': 920,\n",
              " 'एकम': 607,\n",
              " 'वह': 1807,\n",
              " 'थस': 1042,\n",
              " 'जग': 918,\n",
              " 'ययन': 1474,\n",
              " 'उद': 546,\n",
              " 'हरण': 1993,\n",
              " 'थर': 1038,\n",
              " 'शव': 1836,\n",
              " 'भवन': 1321,\n",
              " 'तकन': 1001,\n",
              " 'रब': 1601,\n",
              " 'दन': 1047,\n",
              " 'यन': 1463,\n",
              " 'आकर': 431,\n",
              " 'पन': 1172,\n",
              " 'थम': 1037,\n",
              " 'आपद': 463,\n",
              " 'अवस': 383,\n",
              " 'हतर': 1969,\n",
              " 'सबक': 1897,\n",
              " 'रय': 1658,\n",
              " 'आक': 430,\n",
              " 'अवसर': 384,\n",
              " 'रह इन': 1688,\n",
              " 'सम यक': 1917,\n",
              " 'उनक पर': 565,\n",
              " 'अध ययन': 292,\n",
              " 'उद हरण': 547,\n",
              " 'रद सक': 1587,\n",
              " 'रब धन': 1602,\n",
              " 'सरक कर': 1941,\n",
              " 'कर रत': 720,\n",
              " 'ऐस जह': 651,\n",
              " 'गय गय': 839,\n",
              " 'india': 115,\n",
              " 'अपड': 331,\n",
              " 'हत': 1967,\n",
              " 'मय': 1355,\n",
              " 'jun': 143,\n",
              " 'रध': 1592,\n",
              " 'तय': 1017,\n",
              " 'कत': 692,\n",
              " 'सकर': 1870,\n",
              " 'मन': 1350,\n",
              " 'रन': 1595,\n",
              " 'चर': 895,\n",
              " 'महत': 1447,\n",
              " 'वप': 1787,\n",
              " 'आइस': 403,\n",
              " 'pradesh': 184,\n",
              " 'taking': 211,\n",
              " 'to': 229,\n",
              " 'which': 258,\n",
              " 'are': 15,\n",
              " 'june': 147,\n",
              " 'the': 217,\n",
              " 'patients': 175,\n",
              " 'death': 65,\n",
              " 'toll': 232,\n",
              " 'शख': 1810,\n",
              " 'coronavirus cases': 45,\n",
              " 'cases in': 31,\n",
              " 'in india': 106,\n",
              " 'india updates': 117,\n",
              " 'यरस covid': 1483,\n",
              " 'रह कर': 1698,\n",
              " 'कर हज': 723,\n",
              " 'रम कड': 1610,\n",
              " 'कड पह': 686,\n",
              " 'पह गय': 1214,\n",
              " 'गय लय': 855,\n",
              " 'यरस मर': 1502,\n",
              " 'गई जबक': 785,\n",
              " 'आए गई': 417,\n",
              " 'छल यरस': 912,\n",
              " 'हत मर': 1968,\n",
              " 'कवर रत': 736,\n",
              " 'शत पह': 1814,\n",
              " 'गय jun': 823,\n",
              " 'jun ist': 144,\n",
              " 'बढ पह': 1260,\n",
              " 'पह गए': 1213,\n",
              " ...}"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 309
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Kr636yOnFJVN",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "6771e984-46b6-40e7-d947-19d9f115acc0"
      },
      "source": [
        "len(vectorizer.get_feature_names())"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "2000"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 310
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "qkWgDFV3FYKL",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "463e9656-76b8-48cb-8d7b-f9d09b8db9b8"
      },
      "source": [
        "X.shape"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(843, 2000)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 311
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "9DuGOojpFmMq",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "y = df['Category']"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "alrM9v3HFpDR",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "dd782b94-9a6a-45fc-a9d7-c5371076d8c4"
      },
      "source": [
        "y.shape"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(843,)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 313
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "URA0ZkcVgN7t",
        "colab_type": "text"
      },
      "source": [
        "TRAIN_TEST_SPLIT"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "9D3IeKGgFqHw",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from sklearn.model_selection import train_test_split\n",
        "X_train,X_test,y_train,y_test = train_test_split(X,df['Category'],test_size=0.33,random_state=0)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "2QTo_HRSF0-N",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 238
        },
        "outputId": "958d06ba-056d-4737-84b1-902453e69d54"
      },
      "source": [
        "X_train"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[0.        , 0.        , 0.        , ..., 0.        , 0.        ,\n",
              "        0.        ],\n",
              "       [0.        , 0.        , 0.        , ..., 0.        , 0.        ,\n",
              "        0.        ],\n",
              "       [0.        , 0.        , 0.        , ..., 0.        , 0.        ,\n",
              "        0.        ],\n",
              "       ...,\n",
              "       [0.        , 0.        , 0.        , ..., 0.        , 0.        ,\n",
              "        0.        ],\n",
              "       [0.        , 0.        , 0.        , ..., 0.        , 0.        ,\n",
              "        0.        ],\n",
              "       [0.        , 0.        , 0.15681927, ..., 0.        , 0.        ,\n",
              "        0.        ]])"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 315
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "cKg8CNQ_gcP_",
        "colab_type": "text"
      },
      "source": [
        "# **RUNNING THE MODELS**\n",
        "\n",
        "---\n",
        "\n",
        "\n",
        "\n",
        "---\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "f4WFkU2DhWC2",
        "colab_type": "text"
      },
      "source": [
        "MULTINOMIAL NAIVE_BAYES"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "JYFAruxlF3XO",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "600530f7-1c18-4bb7-8392-569fbc7820dd"
      },
      "source": [
        "from sklearn.naive_bayes import MultinomialNB\n",
        "classifier = MultinomialNB()\n",
        "classifier.fit(X_train,y_train)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "MultinomialNB(alpha=1.0, class_prior=None, fit_prior=True)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 316
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "24_1VhDNGCuy",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 102
        },
        "outputId": "7ef8f131-b1ba-4866-92f6-f93eeddef097"
      },
      "source": [
        "from sklearn.metrics import confusion_matrix\n",
        "from sklearn.metrics import  f1_score,precision_score,recall_score\n",
        "from sklearn.metrics import accuracy_score\n",
        "from sklearn.naive_bayes import MultinomialNB\n",
        "from sklearn.metrics import cohen_kappa_score\n",
        "from sklearn.metrics import roc_auc_score\n",
        "y_pred = classifier.predict(X_test)\n",
        "print(\"ACCURACY \")\n",
        "print(accuracy_score(y_pred,y_test))\n",
        "cm = confusion_matrix(y_test, y_pred)\n",
        "print(\"CONFUSION MATRIX\")\n",
        "print(cm)\n",
        "y_pred = classifier.predict(X_test)\n",
        "acc = accuracy_score(y_test,y_pred)\n",
        "prec = precision_score(y_test,y_pred)\n",
        "rec = recall_score(y_test,y_pred)\n",
        "f1 = f1_score(y_test,y_pred)\n",
        "kappa=cohen_kappa_score(y_test,y_pred)\n",
        "roc=roc_auc_score(y_test,y_pred)\n",
        "results=pd.DataFrame([['MultinomialNB',acc,prec,rec,f1,kappa,roc,cm]],columns=['Model','Accuracy','Precision','Recall','F1 Score','Cohen_kappa_score','ROC','Confusion_matrix'])"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "ACCURACY \n",
            "0.9247311827956989\n",
            "CONFUSION MATRIX\n",
            "[[140   5]\n",
            " [ 16 118]]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "8uguJt-ZGFWn",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 80
        },
        "outputId": "38210829-476d-4610-b95c-b4361f8bda60"
      },
      "source": [
        "results"
      ],
      "execution_count": null,
      "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>Model</th>\n",
              "      <th>Accuracy</th>\n",
              "      <th>Precision</th>\n",
              "      <th>Recall</th>\n",
              "      <th>F1 Score</th>\n",
              "      <th>Cohen_kappa_score</th>\n",
              "      <th>ROC</th>\n",
              "      <th>Confusion_matrix</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>MultinomialNB</td>\n",
              "      <td>0.924731</td>\n",
              "      <td>0.95935</td>\n",
              "      <td>0.880597</td>\n",
              "      <td>0.918288</td>\n",
              "      <td>0.848757</td>\n",
              "      <td>0.923057</td>\n",
              "      <td>[[140, 5], [16, 118]]</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "           Model  Accuracy  ...       ROC       Confusion_matrix\n",
              "0  MultinomialNB  0.924731  ...  0.923057  [[140, 5], [16, 118]]\n",
              "\n",
              "[1 rows x 8 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 318
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "54ipu8VriyS9",
        "colab_type": "text"
      },
      "source": [
        "LOGISTIC REGRESSION"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "uhv8CYyaGI9Z",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 102
        },
        "outputId": "b6b7625d-8ac7-418e-9921-e0225791dcaa"
      },
      "source": [
        "from sklearn.linear_model import LogisticRegression\n",
        "from sklearn.naive_bayes import MultinomialNB\n",
        "logreg = LogisticRegression(C=1e5)\n",
        "logreg.fit(X_train, y_train)\n",
        "y_pred = logreg.predict(X_test)\n",
        "print('Accuracy  on training set: {:.2f}'.format(classifier.score(X_train, y_train)))\n",
        "print('Accuracy on test set: {:.2f}'.format(classifier.score(X_test, y_test)))\n",
        "cm = confusion_matrix(y_test, y_pred)\n",
        "print(\"CONFUSION MATRIX\")\n",
        "print(cm)\n",
        "acc = accuracy_score(y_test,y_pred)\n",
        "prec = precision_score(y_test,y_pred)\n",
        "rec = recall_score(y_test,y_pred)\n",
        "f1 = f1_score(y_test,y_pred)\n",
        "kappa=cohen_kappa_score(y_test,y_pred)\n",
        "roc=roc_auc_score(y_test,y_pred)\n",
        "model_results=pd.DataFrame([['Logistic Regression',acc,prec,rec,f1,kappa,roc,cm]],columns=['Model','Accuracy','Precision','Recall','F1 Score','Cohen_kappa_score','ROC','Confusion_matrix'])\n",
        "results = results.append(model_results,ignore_index=True)\n"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Accuracy  on training set: 0.98\n",
            "Accuracy on test set: 0.92\n",
            "CONFUSION MATRIX\n",
            "[[141   4]\n",
            " [  4 130]]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "R_kd1Qgij1iA",
        "colab_type": "text"
      },
      "source": [
        "SVM-rbf kernel"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "lOAzJDx2GSj8",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 85
        },
        "outputId": "6560d421-8736-46a9-cdbc-ec65c7ff7f3b"
      },
      "source": [
        "from sklearn.svm import SVC\n",
        "classifier = SVC()\n",
        "classifier.fit(X_train,y_train)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "SVC(C=1.0, break_ties=False, cache_size=200, class_weight=None, coef0=0.0,\n",
              "    decision_function_shape='ovr', degree=3, gamma='scale', kernel='rbf',\n",
              "    max_iter=-1, probability=False, random_state=None, shrinking=True,\n",
              "    tol=0.001, verbose=False)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 320
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "hWvyGmDkGadE",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 102
        },
        "outputId": "35f93990-58b8-4516-e1d5-40d49ba4c069"
      },
      "source": [
        "y_pred = classifier.predict(X_test)\n",
        "print('Accuracy  on training set: {:.2f}'.format(classifier.score(X_train, y_train)))\n",
        "print('Accuracy on test set: {:.2f}'.format(classifier.score(X_test, y_test)))\n",
        "cm = confusion_matrix(y_test, y_pred)\n",
        "print(\"CONFUSION MATRIX\")\n",
        "print(cm)\n",
        "acc = accuracy_score(y_test,y_pred)\n",
        "prec = precision_score(y_test,y_pred)\n",
        "rec = recall_score(y_test,y_pred)\n",
        "f1 = f1_score(y_test,y_pred)\n",
        "kappa=cohen_kappa_score(y_test,y_pred)\n",
        "roc=roc_auc_score(y_test,y_pred)\n",
        "model_results=pd.DataFrame([['SVM RBF',acc,prec,rec,f1,kappa,roc,cm]],columns=['Model','Accuracy','Precision','Recall','F1 Score','Cohen_kappa_score','ROC','Confusion_matrix'])\n",
        "results = results.append(model_results,ignore_index=True)\n"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Accuracy  on training set: 1.00\n",
            "Accuracy on test set: 0.97\n",
            "CONFUSION MATRIX\n",
            "[[143   2]\n",
            " [  6 128]]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "AAMWccHRGcTG",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 142
        },
        "outputId": "ea67bd9a-6b0f-4c52-d9a3-792c8d1196b8"
      },
      "source": [
        "results"
      ],
      "execution_count": null,
      "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>Model</th>\n",
              "      <th>Accuracy</th>\n",
              "      <th>Precision</th>\n",
              "      <th>Recall</th>\n",
              "      <th>F1 Score</th>\n",
              "      <th>Cohen_kappa_score</th>\n",
              "      <th>ROC</th>\n",
              "      <th>Confusion_matrix</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>MultinomialNB</td>\n",
              "      <td>0.924731</td>\n",
              "      <td>0.959350</td>\n",
              "      <td>0.880597</td>\n",
              "      <td>0.918288</td>\n",
              "      <td>0.848757</td>\n",
              "      <td>0.923057</td>\n",
              "      <td>[[140, 5], [16, 118]]</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>Logistic Regression</td>\n",
              "      <td>0.971326</td>\n",
              "      <td>0.970149</td>\n",
              "      <td>0.970149</td>\n",
              "      <td>0.970149</td>\n",
              "      <td>0.942563</td>\n",
              "      <td>0.971282</td>\n",
              "      <td>[[141, 4], [4, 130]]</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>SVM RBF</td>\n",
              "      <td>0.971326</td>\n",
              "      <td>0.984615</td>\n",
              "      <td>0.955224</td>\n",
              "      <td>0.969697</td>\n",
              "      <td>0.942498</td>\n",
              "      <td>0.970715</td>\n",
              "      <td>[[143, 2], [6, 128]]</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "                 Model  Accuracy  ...       ROC       Confusion_matrix\n",
              "0        MultinomialNB  0.924731  ...  0.923057  [[140, 5], [16, 118]]\n",
              "1  Logistic Regression  0.971326  ...  0.971282   [[141, 4], [4, 130]]\n",
              "2              SVM RBF  0.971326  ...  0.970715   [[143, 2], [6, 128]]\n",
              "\n",
              "[3 rows x 8 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 322
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "h4gBREpFlxXT",
        "colab_type": "text"
      },
      "source": [
        "SVM-linear kernel"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "piWbODqDGeSV",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 85
        },
        "outputId": "d9c8355f-5765-4f10-8901-6e19178ebc0c"
      },
      "source": [
        "from sklearn.svm import SVC\n",
        "classifier = SVC(kernel=\"linear\")\n",
        "classifier.fit(X_train,y_train)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "SVC(C=1.0, break_ties=False, cache_size=200, class_weight=None, coef0=0.0,\n",
              "    decision_function_shape='ovr', degree=3, gamma='scale', kernel='linear',\n",
              "    max_iter=-1, probability=False, random_state=None, shrinking=True,\n",
              "    tol=0.001, verbose=False)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 323
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "jGBxr05_GjA1",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 102
        },
        "outputId": "b6172335-b3dc-4d61-ed90-92ce994bdf55"
      },
      "source": [
        "y_pred = classifier.predict(X_test)\n",
        "print('Accuracy  on training set: {:.2f}'.format(classifier.score(X_train, y_train)))\n",
        "print('Accuracy on test set: {:.2f}'.format(classifier.score(X_test, y_test)))\n",
        "cm = confusion_matrix(y_test, y_pred)\n",
        "print(\"CONFUSION MATRIX\")\n",
        "print(cm)\n",
        "acc = accuracy_score(y_test,y_pred)\n",
        "prec = precision_score(y_test,y_pred)\n",
        "rec = recall_score(y_test,y_pred)\n",
        "f1 = f1_score(y_test,y_pred)\n",
        "kappa=cohen_kappa_score(y_test,y_pred)\n",
        "roc=roc_auc_score(y_test,y_pred)\n",
        "model_results=pd.DataFrame([['SVM linear',acc,prec,rec,f1,kappa,roc,cm]],columns=['Model','Accuracy','Precision','Recall','F1 Score','Cohen_kappa_score','ROC','Confusion_matrix'])\n",
        "results = results.append(model_results,ignore_index=True)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Accuracy  on training set: 0.99\n",
            "Accuracy on test set: 0.97\n",
            "CONFUSION MATRIX\n",
            "[[141   4]\n",
            " [  4 130]]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "LhlMS0sCGofY",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 173
        },
        "outputId": "c6a76dd1-28d9-4964-f6b0-5db29ac4daf2"
      },
      "source": [
        "results"
      ],
      "execution_count": null,
      "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>Model</th>\n",
              "      <th>Accuracy</th>\n",
              "      <th>Precision</th>\n",
              "      <th>Recall</th>\n",
              "      <th>F1 Score</th>\n",
              "      <th>Cohen_kappa_score</th>\n",
              "      <th>ROC</th>\n",
              "      <th>Confusion_matrix</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>MultinomialNB</td>\n",
              "      <td>0.924731</td>\n",
              "      <td>0.959350</td>\n",
              "      <td>0.880597</td>\n",
              "      <td>0.918288</td>\n",
              "      <td>0.848757</td>\n",
              "      <td>0.923057</td>\n",
              "      <td>[[140, 5], [16, 118]]</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>Logistic Regression</td>\n",
              "      <td>0.971326</td>\n",
              "      <td>0.970149</td>\n",
              "      <td>0.970149</td>\n",
              "      <td>0.970149</td>\n",
              "      <td>0.942563</td>\n",
              "      <td>0.971282</td>\n",
              "      <td>[[141, 4], [4, 130]]</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>SVM RBF</td>\n",
              "      <td>0.971326</td>\n",
              "      <td>0.984615</td>\n",
              "      <td>0.955224</td>\n",
              "      <td>0.969697</td>\n",
              "      <td>0.942498</td>\n",
              "      <td>0.970715</td>\n",
              "      <td>[[143, 2], [6, 128]]</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>SVM linear</td>\n",
              "      <td>0.971326</td>\n",
              "      <td>0.970149</td>\n",
              "      <td>0.970149</td>\n",
              "      <td>0.970149</td>\n",
              "      <td>0.942563</td>\n",
              "      <td>0.971282</td>\n",
              "      <td>[[141, 4], [4, 130]]</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "                 Model  Accuracy  ...       ROC       Confusion_matrix\n",
              "0        MultinomialNB  0.924731  ...  0.923057  [[140, 5], [16, 118]]\n",
              "1  Logistic Regression  0.971326  ...  0.971282   [[141, 4], [4, 130]]\n",
              "2              SVM RBF  0.971326  ...  0.970715   [[143, 2], [6, 128]]\n",
              "3           SVM linear  0.971326  ...  0.971282   [[141, 4], [4, 130]]\n",
              "\n",
              "[4 rows x 8 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 325
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Zhgbl3CymETk",
        "colab_type": "text"
      },
      "source": [
        "PASSIVE AGGRESIVE CLASSIFIER"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "23oDe119GrYH",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 119
        },
        "outputId": "a64b952b-ecd2-4b15-a8ad-aaf9c2cb8978"
      },
      "source": [
        "from sklearn.linear_model import PassiveAggressiveClassifier\n",
        "classifier = PassiveAggressiveClassifier()\n",
        "classifier.fit(X_train,y_train)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "PassiveAggressiveClassifier(C=1.0, average=False, class_weight=None,\n",
              "                            early_stopping=False, fit_intercept=True,\n",
              "                            loss='hinge', max_iter=1000, n_iter_no_change=5,\n",
              "                            n_jobs=None, random_state=None, shuffle=True,\n",
              "                            tol=0.001, validation_fraction=0.1, verbose=0,\n",
              "                            warm_start=False)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 326
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "vAEHv_lDGwlh",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 102
        },
        "outputId": "f4f57c81-1a20-41ae-939f-f173e236265a"
      },
      "source": [
        "y_pred = classifier.predict(X_test)\n",
        "print('Accuracy  on training set: {:.2f}'.format(classifier.score(X_train, y_train)))\n",
        "print('Accuracy on test set: {:.2f}'.format(classifier.score(X_test, y_test)))\n",
        "cm = confusion_matrix(y_test, y_pred)\n",
        "print(\"CONFUSION MATRIX\")\n",
        "print(cm)\n",
        "acc = accuracy_score(y_test,y_pred)\n",
        "prec = precision_score(y_test,y_pred)\n",
        "rec = recall_score(y_test,y_pred)\n",
        "f1 = f1_score(y_test,y_pred)\n",
        "kappa=cohen_kappa_score(y_test,y_pred)\n",
        "roc=roc_auc_score(y_test,y_pred)\n",
        "model_results=pd.DataFrame([['Passive Aggresive',acc,prec,rec,f1,kappa,roc,cm]],columns=['Model','Accuracy','Precision','Recall','F1 Score','Cohen_kappa_score','ROC','Confusion_matrix'])\n",
        "results = results.append(model_results,ignore_index=True)\n",
        "\n"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Accuracy  on training set: 1.00\n",
            "Accuracy on test set: 0.97\n",
            "CONFUSION MATRIX\n",
            "[[141   4]\n",
            " [  4 130]]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "csjvcB3ameAY",
        "colab_type": "text"
      },
      "source": [
        "ANN"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "kn_rEqN4GyM_",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 207
        },
        "outputId": "c6142fb6-34ac-48ab-9f2c-9c673974e172"
      },
      "source": [
        "import tensorflow as tf\n",
        "from tensorflow.keras.callbacks import EarlyStopping\n",
        "from tensorflow.keras.optimizers import Adam\n",
        "early_stop = EarlyStopping(monitor='val_loss',patience=1.5)\n",
        "N, D = X_train.shape\n",
        "model = tf.keras.models.Sequential([\n",
        "  tf.keras.layers.Input(shape=(D,)),\n",
        "  tf.keras.layers.Dense(300,activation='relu'),\n",
        "  tf.keras.layers.Dropout(0.7),\n",
        "  tf.keras.layers.Dense(50,activation='relu'),\n",
        "  tf.keras.layers.Dropout(0.7),\n",
        "  #tf.keras.layers.Dense(50,activation='relu'),\n",
        "  #tf.keras.layers.Dropout(0.7),\n",
        "  #tf.keras.layers.Dense(50,activation='relu'),\n",
        "  #tf.keras.layers.Dense(10,activation='relu'),\n",
        "  tf.keras.layers.Dense(1, activation='sigmoid')\n",
        "])\n",
        "model.compile(Adam(lr=0.005),\n",
        "              loss='binary_crossentropy',\n",
        "              metrics=['accuracy'])\n",
        "his = model.fit(X_train,y_train,epochs=25,verbose=1,shuffle=1,validation_split=0.1,callbacks=[early_stop])"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/25\n",
            "16/16 [==============================] - 0s 13ms/step - loss: 0.6157 - accuracy: 0.6746 - val_loss: 0.3864 - val_accuracy: 1.0000\n",
            "Epoch 2/25\n",
            "16/16 [==============================] - 0s 8ms/step - loss: 0.2361 - accuracy: 0.9467 - val_loss: 0.0388 - val_accuracy: 1.0000\n",
            "Epoch 3/25\n",
            "16/16 [==============================] - 0s 10ms/step - loss: 0.0607 - accuracy: 0.9842 - val_loss: 0.0102 - val_accuracy: 1.0000\n",
            "Epoch 4/25\n",
            "16/16 [==============================] - 0s 10ms/step - loss: 0.0382 - accuracy: 0.9882 - val_loss: 0.0173 - val_accuracy: 1.0000\n",
            "Epoch 5/25\n",
            "16/16 [==============================] - 0s 9ms/step - loss: 0.0306 - accuracy: 0.9901 - val_loss: 0.0288 - val_accuracy: 0.9825\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "mHgTiXkFG3x8",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 279
        },
        "outputId": "2ceca2ae-f982-4114-c502-4ea1fce36ae9"
      },
      "source": [
        "import matplotlib.pyplot as plt\n",
        "plt.plot(his.history['loss'])\n",
        "plt.plot(his.history['val_loss'])\n",
        "plt.xlabel(\"Epochs\")\n",
        "plt.legend([\"Loss\",\"Val_loss\"])\n",
        "plt.ylabel(\"Loss\")\n",
        "plt.savefig('loss_plot.png', dpi=300, bbox_inches='tight')\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Sa_jq53sG6hC",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 265
        },
        "outputId": "945cfbd9-24eb-4ea2-8774-2d3a3365c311"
      },
      "source": [
        "plt.plot(his.history['accuracy'], label='train')\n",
        "plt.plot(his.history['val_accuracy'], label='test')\n",
        "\n",
        "plt.legend([\"Accuracy\",\"Val_accuracy\"])\n",
        "plt.savefig('accuracy_plot.png', dpi=300, bbox_inches='tight')\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "YZznKriQG-jQ",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "results.to_csv(\"finalresults2.csv\")"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "x0o00jVLHBjZ",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 153
        },
        "outputId": "2a271702-810e-4694-91e6-43c9b9411f1d"
      },
      "source": [
        "# demonstration of calculating metrics for a neural network model using sklearn\n",
        "from sklearn.metrics import accuracy_score\n",
        "from sklearn.metrics import precision_score\n",
        "from sklearn.metrics import recall_score\n",
        "from sklearn.metrics import f1_score\n",
        "from sklearn.metrics import cohen_kappa_score\n",
        "from sklearn.metrics import roc_auc_score\n",
        "from sklearn.metrics import confusion_matrix\n",
        " \n",
        "trainX, trainy, testX, testy = X_train,y_train,X_test,y_test \n",
        "yhat_probs = (model.predict(testX) > 0.5).astype(\"int32\")\n",
        "# predict crisp classes for test set\n",
        "yhat_classes = model.predict_classes(testX, verbose=0)\n",
        "# reduce to 1d array\n",
        "yhat_probs = yhat_probs[:, 0]\n",
        "yhat_classes = yhat_classes[:, 0]\n",
        " \n",
        "# accuracy: (tp + tn) / (p + n)\n",
        "accuracy = accuracy_score(testy, yhat_classes)\n",
        "print('Accuracy: %f' % accuracy)\n",
        "# precision tp / (tp + fp)\n",
        "precision = precision_score(testy, yhat_classes)\n",
        "print('Precision: %f' % precision)\n",
        "# recall: tp / (tp + fn)\n",
        "recall = recall_score(testy, yhat_classes)\n",
        "print('Recall: %f' % recall)\n",
        "# f1: 2 tp / (2 tp + fp + fn)\n",
        "f1 = f1_score(testy, yhat_classes)\n",
        "print('F1 score: %f' % f1)\n",
        " \n",
        "# kappa\n",
        "kappa = cohen_kappa_score(testy, yhat_classes)\n",
        "print('Cohens kappa: %f' % kappa)\n",
        "# ROC AUC\n",
        "auc = roc_auc_score(testy, yhat_probs)\n",
        "print('ROC AUC: %f' % auc)\n",
        "# confusion matrix\n",
        "matrix = confusion_matrix(testy, yhat_classes)\n",
        "print(matrix)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Accuracy: 0.946237\n",
            "Precision: 0.916084\n",
            "Recall: 0.977612\n",
            "F1 score: 0.945848\n",
            "Cohens kappa: 0.892579\n",
            "ROC AUC: 0.947427\n",
            "[[133  12]\n",
            " [  3 131]]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ckXSBPSLHEke",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "model_results=pd.DataFrame([['ANN2',accuracy,precision,recall,f1,kappa,auc,matrix]],columns=['Model','Accuracy','Precision','Recall','F1 Score','Cohen_kappa_score','ROC','Confusion_matrix'])\n",
        "results = results.append(model_results,ignore_index=True)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "8QO__irrHMz_",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 235
        },
        "outputId": "3759f242-9f64-413b-85cb-bddb4b86095d"
      },
      "source": [
        "results"
      ],
      "execution_count": null,
      "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>Model</th>\n",
              "      <th>Accuracy</th>\n",
              "      <th>Precision</th>\n",
              "      <th>Recall</th>\n",
              "      <th>F1 Score</th>\n",
              "      <th>Cohen_kappa_score</th>\n",
              "      <th>ROC</th>\n",
              "      <th>Confusion_matrix</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>MultinomialNB</td>\n",
              "      <td>0.924731</td>\n",
              "      <td>0.959350</td>\n",
              "      <td>0.880597</td>\n",
              "      <td>0.918288</td>\n",
              "      <td>0.848757</td>\n",
              "      <td>0.923057</td>\n",
              "      <td>[[140, 5], [16, 118]]</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>Logistic Regression</td>\n",
              "      <td>0.971326</td>\n",
              "      <td>0.970149</td>\n",
              "      <td>0.970149</td>\n",
              "      <td>0.970149</td>\n",
              "      <td>0.942563</td>\n",
              "      <td>0.971282</td>\n",
              "      <td>[[141, 4], [4, 130]]</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>SVM RBF</td>\n",
              "      <td>0.971326</td>\n",
              "      <td>0.984615</td>\n",
              "      <td>0.955224</td>\n",
              "      <td>0.969697</td>\n",
              "      <td>0.942498</td>\n",
              "      <td>0.970715</td>\n",
              "      <td>[[143, 2], [6, 128]]</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>SVM linear</td>\n",
              "      <td>0.971326</td>\n",
              "      <td>0.970149</td>\n",
              "      <td>0.970149</td>\n",
              "      <td>0.970149</td>\n",
              "      <td>0.942563</td>\n",
              "      <td>0.971282</td>\n",
              "      <td>[[141, 4], [4, 130]]</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>Passive Aggresive</td>\n",
              "      <td>0.971326</td>\n",
              "      <td>0.970149</td>\n",
              "      <td>0.970149</td>\n",
              "      <td>0.970149</td>\n",
              "      <td>0.942563</td>\n",
              "      <td>0.971282</td>\n",
              "      <td>[[141, 4], [4, 130]]</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>ANN2</td>\n",
              "      <td>0.946237</td>\n",
              "      <td>0.916084</td>\n",
              "      <td>0.977612</td>\n",
              "      <td>0.945848</td>\n",
              "      <td>0.892579</td>\n",
              "      <td>0.947427</td>\n",
              "      <td>[[133, 12], [3, 131]]</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "                 Model  Accuracy  ...       ROC       Confusion_matrix\n",
              "0        MultinomialNB  0.924731  ...  0.923057  [[140, 5], [16, 118]]\n",
              "1  Logistic Regression  0.971326  ...  0.971282   [[141, 4], [4, 130]]\n",
              "2              SVM RBF  0.971326  ...  0.970715   [[143, 2], [6, 128]]\n",
              "3           SVM linear  0.971326  ...  0.971282   [[141, 4], [4, 130]]\n",
              "4    Passive Aggresive  0.971326  ...  0.971282   [[141, 4], [4, 130]]\n",
              "5                 ANN2  0.946237  ...  0.947427  [[133, 12], [3, 131]]\n",
              "\n",
              "[6 rows x 8 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 334
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "IsbE5JtgHOgW",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "b124c3d8-d700-441a-dc95-4ed6d7bcbbff"
      },
      "source": [
        "def build_corpus(data):\n",
        "    \"Creates a list of lists containing words from each sentence\"\n",
        "    corpus = []\n",
        "    for col in data:\n",
        "        \n",
        "            corpus.append(col.split())\n",
        "\n",
        "    return corpus\n",
        "\n",
        "corpus1 = build_corpus(df['stopwords'])\n",
        "corpus1[0:2]"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[['सोनभद्र',\n",
              "  'जिल',\n",
              "  'सोमवार',\n",
              "  'लोग',\n",
              "  'जांच',\n",
              "  'रिपोर्ट',\n",
              "  'कोरो',\n",
              "  'पाजिटिव',\n",
              "  'आई',\n",
              "  'है।',\n",
              "  'इनम',\n",
              "  'संक्रमित',\n",
              "  'गुरम',\n",
              "  'स्थित',\n",
              "  'जिल',\n",
              "  'जेल',\n",
              "  'परिसर',\n",
              "  'जुड़',\n",
              "  'हैं।',\n",
              "  'इनम',\n",
              "  'आठ',\n",
              "  'जेल',\n",
              "  'कर्म',\n",
              "  'नौ',\n",
              "  'बंद',\n",
              "  'शामिल',\n",
              "  'हैं।',\n",
              "  'इनक',\n",
              "  'अलाव',\n",
              "  'तीन',\n",
              "  'लोग',\n",
              "  'राबर्ट्सगंज',\n",
              "  'बरवन',\n",
              "  'लोहर',\n",
              "  'न्य',\n",
              "  'मार्केट',\n",
              "  'रामलील',\n",
              "  'मैदान',\n",
              "  'पास',\n",
              "  'रह',\n",
              "  'वाल',\n",
              "  'हैं।',\n",
              "  'इसक',\n",
              "  'पुष्ट',\n",
              "  'सीएमओ',\n",
              "  'डा',\n",
              "  'एसक',\n",
              "  'उपाध्याय',\n",
              "  'है।',\n",
              "  'जिल',\n",
              "  'दिन',\n",
              "  'अब',\n",
              "  'सबस',\n",
              "  'ज्याद',\n",
              "  'आठ',\n",
              "  'लोग',\n",
              "  'संक्त्रस्मित',\n",
              "  'मिल',\n",
              "  'थे।',\n",
              "  'इसक',\n",
              "  'सोमवार',\n",
              "  'इत',\n",
              "  'मरीज',\n",
              "  'मिले।',\n",
              "  'इसक',\n",
              "  'संक्रमित',\n",
              "  'संख्य',\n",
              "  'गई।',\n",
              "  'इनम',\n",
              "  'लोग',\n",
              "  'स्वस्थ',\n",
              "  'लौट',\n",
              "  'गए',\n",
              "  'जबक',\n",
              "  'लोग',\n",
              "  'मधुपुर',\n",
              "  'कोविड',\n",
              "  'केयर',\n",
              "  'सेंटर',\n",
              "  'इलाज',\n",
              "  'चल',\n",
              "  'रह',\n",
              "  'है।'],\n",
              " ['Image',\n",
              "  'Source',\n",
              "  'INSTAGRAMTANUSRIDASGUPTA',\n",
              "  'बालाज',\n",
              "  'टेलीफिल्म्स',\n",
              "  'क्रिएटिव',\n",
              "  'हेड',\n",
              "  'कोरो',\n",
              "  'समय',\n",
              "  'पूर',\n",
              "  'देश',\n",
              "  'कोरो',\n",
              "  'वायरस',\n",
              "  'जूझ',\n",
              "  'रह',\n",
              "  'है।',\n",
              "  'अब',\n",
              "  'फिल्म',\n",
              "  'टीव',\n",
              "  'इंडस्ट्र',\n",
              "  'दस्तक',\n",
              "  'दे',\n",
              "  'दी',\n",
              "  'है।',\n",
              "  'बॉलीवुड',\n",
              "  'महानायक',\n",
              "  'अमिताभ',\n",
              "  'बच्चन',\n",
              "  'अभिषेक',\n",
              "  'ऐश्वर्य',\n",
              "  'आराध्य',\n",
              "  'कोरो',\n",
              "  'संक्रमित',\n",
              "  'गए।',\n",
              "  'अनुपम',\n",
              "  'खेर',\n",
              "  'फैमिल',\n",
              "  'इसक',\n",
              "  'शिकार',\n",
              "  'गई।',\n",
              "  'टीव',\n",
              "  'एक्टर',\n",
              "  'पार्थ',\n",
              "  'समथान',\n",
              "  'कोविड',\n",
              "  'गया।',\n",
              "  'अब',\n",
              "  'कपूर',\n",
              "  'बालाज',\n",
              "  'टेलीफिल्म्स',\n",
              "  'क्रिएटिव',\n",
              "  'हेड',\n",
              "  'तनुश्र',\n",
              "  'दासगुप्',\n",
              "  'इसक',\n",
              "  'चपेट',\n",
              "  'आ',\n",
              "  'गई',\n",
              "  'हैं।',\n",
              "  'तनुश्र',\n",
              "  'दासगुप्',\n",
              "  'बत',\n",
              "  'वो',\n",
              "  'पहल',\n",
              "  'होम',\n",
              "  'क्वारंटीन',\n",
              "  'थीं',\n",
              "  'जैस',\n",
              "  'ऑक्सीजन',\n",
              "  'लेवल',\n",
              "  'गिर',\n",
              "  'लग',\n",
              "  'उन्ह',\n",
              "  'जुल',\n",
              "  'मुंबई',\n",
              "  'सेवन',\n",
              "  'हिल्स',\n",
              "  'हॉस्पिटल',\n",
              "  'एडमिट',\n",
              "  'गया।',\n",
              "  'उनक',\n",
              "  'मां',\n",
              "  'सबस',\n",
              "  'पहल',\n",
              "  'कोरो',\n",
              "  'था।',\n",
              "  'उन्ह',\n",
              "  'माइल्ड',\n",
              "  'लक्षण',\n",
              "  'थे।',\n",
              "  'पार्थ',\n",
              "  'समथान',\n",
              "  'कोरो',\n",
              "  'पॉजिटिव',\n",
              "  'कसौट',\n",
              "  'जिंदग',\n",
              "  'नए',\n",
              "  'एपिसोड',\n",
              "  'नह',\n",
              "  'देख',\n",
              "  'पा',\n",
              "  'इंडिय',\n",
              "  'फोरम्स',\n",
              "  'इंटरव्य',\n",
              "  'दे',\n",
              "  'तनुश्र',\n",
              "  'कह',\n",
              "  'ऑक्सीजन',\n",
              "  'लेवल',\n",
              "  'कम',\n",
              "  'मुझ',\n",
              "  'शनिवार',\n",
              "  'अस्पताल',\n",
              "  'लाय',\n",
              "  'गया।',\n",
              "  'डॉक्टर',\n",
              "  'ब्लड',\n",
              "  'टेस्ट',\n",
              "  'किय',\n",
              "  'मुझ',\n",
              "  'एडमिट',\n",
              "  'सलाह',\n",
              "  'दी',\n",
              "  'ताक',\n",
              "  'मेर',\n",
              "  'निगरा',\n",
              "  'सके।',\n",
              "  'इसस',\n",
              "  'पहल',\n",
              "  'मैं',\n",
              "  'अप',\n",
              "  'मां',\n",
              "  'खुद',\n",
              "  'कोरो',\n",
              "  'संक्रमित',\n",
              "  'उन्ह',\n",
              "  'हॉस्पिटल',\n",
              "  'एडमिट',\n",
              "  'जरूरत',\n",
              "  'नह',\n",
              "  'है।',\n",
              "  'वो',\n",
              "  'होम',\n",
              "  'क्वारंटीन',\n",
              "  'धीरेधीर',\n",
              "  'ठीक',\n",
              "  'रह',\n",
              "  'हैं।',\n",
              "  'आग',\n",
              "  'कह',\n",
              "  'मुझ',\n",
              "  'गहर',\n",
              "  'स्ट्रेन',\n",
              "  'जिसक',\n",
              "  'भर्',\n",
              "  'पड़ा।',\n",
              "  'कर्मचार',\n",
              "  'मुझ',\n",
              "  'ठीक',\n",
              "  'मदद',\n",
              "  'रह',\n",
              "  'हैं।',\n",
              "  'मुझ',\n",
              "  'अभ',\n",
              "  'खांस',\n",
              "  'उसक',\n",
              "  'इलाज',\n",
              "  'चल',\n",
              "  'रह',\n",
              "  'है।',\n",
              "  'हां',\n",
              "  'दिन',\n",
              "  'ठीक',\n",
              "  'अगर',\n",
              "  'सह',\n",
              "  'हाथ',\n",
              "  'डर',\n",
              "  'जरूरत',\n",
              "  'नह',\n",
              "  'है।',\n",
              "  'हम',\n",
              "  'सभ',\n",
              "  'ठीक',\n",
              "  'रह',\n",
              "  'मैं',\n",
              "  'जल्द',\n",
              "  'वापस',\n",
              "  'जाऊंगी।',\n",
              "  'कसौट',\n",
              "  'जिंदग',\n",
              "  'एक्टर',\n",
              "  'पार्थ',\n",
              "  'समथान',\n",
              "  'कोरो',\n",
              "  'पॉजिटिव',\n",
              "  'किय',\n",
              "  'पोस्ट',\n",
              "  'गौरतलब',\n",
              "  'इसस',\n",
              "  'पहल',\n",
              "  'कपूर',\n",
              "  'सीरियल',\n",
              "  'कसौट',\n",
              "  'जिंदग',\n",
              "  'फेम',\n",
              "  'पार्थ',\n",
              "  'समथान',\n",
              "  'कोरो',\n",
              "  'संक्रमित',\n",
              "  'गए।',\n",
              "  'पार्थ',\n",
              "  'सीरियल',\n",
              "  'अनुराग',\n",
              "  'लीड',\n",
              "  'रोल',\n",
              "  'निभ',\n",
              "  'हैं।',\n",
              "  'पार्थ',\n",
              "  'रिपोर्ट',\n",
              "  'पॉजिटिव',\n",
              "  'आन',\n",
              "  'शो',\n",
              "  'शूटिंग',\n",
              "  'बंद',\n",
              "  'दी',\n",
              "  'गई',\n",
              "  'है।',\n",
              "  'पार्थ',\n",
              "  'समथान',\n",
              "  'इंस्टाग्राम',\n",
              "  'अप',\n",
              "  'तस्वीर',\n",
              "  'शेयर',\n",
              "  'करक',\n",
              "  'लिख',\n",
              "  'सभ',\n",
              "  'हेल',\n",
              "  'मुझ',\n",
              "  'हल्क',\n",
              "  'लक्षण',\n",
              "  'दिख',\n",
              "  'जिसक',\n",
              "  'मैं',\n",
              "  'अप',\n",
              "  'टेस्ट',\n",
              "  'हां',\n",
              "  'मैं',\n",
              "  'कोविड',\n",
              "  'पॉजिटिव',\n",
              "  'पाय',\n",
              "  'गय',\n",
              "  'हूं।',\n",
              "  'मैं',\n",
              "  'सभ',\n",
              "  'गुजारिश',\n",
              "  'हूं',\n",
              "  'पिछल',\n",
              "  'दिन',\n",
              "  'मेर',\n",
              "  'संपर्क',\n",
              "  'आय',\n",
              "  'अप',\n",
              "  'टेस्ट',\n",
              "  'जरूर',\n",
              "  'कराए।',\n",
              "  'बीएमस',\n",
              "  'लगातार',\n",
              "  'हमार',\n",
              "  'टत',\n",
              "  'डॉक्टर्स',\n",
              "  'गाइडलाइन्स',\n",
              "  'मैं',\n",
              "  'सेल्फ',\n",
              "  'क्वारंटीन',\n",
              "  'हूं।',\n",
              "  'उनक',\n",
              "  'सपोर्ट',\n",
              "  'मैं',\n",
              "  'ग्रेटफुल',\n",
              "  'हूं।',\n",
              "  'प्लीज',\n",
              "  'सभ',\n",
              "  'लोग',\n",
              "  'अप',\n",
              "  'ख्याल',\n",
              "  'रखिएगा।',\n",
              "  'कोरो',\n",
              "  'जंग',\n",
              "  'Full',\n",
              "  'Coverage']]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 335
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "fQeLVhdHHVbQ",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "2d732ae4-0b78-4754-dc58-c5f6dec0c440"
      },
      "source": [
        "from gensim.models import word2vec\n",
        "\n",
        "from sklearn.manifold import TSNE\n",
        "#model = word2vec.Word2Vec(corpus1, size=100, window=20, min_count=200, workers=4)\n",
        "model = word2vec.Word2Vec(corpus1, size=100, window=20, min_count=100, workers=4)\n",
        "def tsne_plot(model):\n",
        "    \"Creates and TSNE model and plots it\"\n",
        "    labels = []\n",
        "    tokens = []\n",
        "\n",
        "    for word in model.wv.vocab:\n",
        "        tokens.append(model[word])\n",
        "        labels.append(word)\n",
        "\n",
        "    tsne_model = TSNE(perplexity=500, n_components=2, init='pca', n_iter=2500, random_state=23)\n",
        "    new_values = tsne_model.fit_transform(tokens)\n",
        "\n",
        "    x = []\n",
        "    y = []\n",
        "    for value in new_values:\n",
        "        x.append(value[0])\n",
        "        y.append(value[1])\n",
        "\n",
        "    plt.figure(figsize=(16, 16)) \n",
        "    for i in range(len(x)):\n",
        "        plt.scatter(x[i],y[i])\n",
        "        plt.annotate(labels[i],\n",
        "                     xy=(x[i], y[i]),\n",
        "                     xytext=(5, 2),\n",
        "                     textcoords='offset points',\n",
        "                     ha='right',\n",
        "                     va='bottom')\n",
        "    plt.savefig('bow1.png', dpi=300, bbox_inches='tight')\n",
        "    plt.show()\n",
        "tsne_plot(model)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/ipykernel_launcher.py:12: DeprecationWarning: Call to deprecated `__getitem__` (Method will be removed in 4.0.0, use self.wv.__getitem__() instead).\n",
            "  if sys.path[0] == '':\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:214: RuntimeWarning: Glyph 2332 missing from current font.\n",
            "  font.set_text(s, 0.0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:214: RuntimeWarning: Glyph 2367 missing from current font.\n",
            "  font.set_text(s, 0.0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:214: RuntimeWarning: Glyph 2354 missing from current font.\n",
            "  font.set_text(s, 0.0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:214: RuntimeWarning: Glyph 2360 missing from current font.\n",
            "  font.set_text(s, 0.0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:214: RuntimeWarning: Glyph 2379 missing from current font.\n",
            "  font.set_text(s, 0.0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:214: RuntimeWarning: Glyph 2350 missing from current font.\n",
            "  font.set_text(s, 0.0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:214: RuntimeWarning: Glyph 2357 missing from current font.\n",
            "  font.set_text(s, 0.0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:214: RuntimeWarning: Glyph 2366 missing from current font.\n",
            "  font.set_text(s, 0.0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:214: RuntimeWarning: Glyph 2352 missing from current font.\n",
            "  font.set_text(s, 0.0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:214: RuntimeWarning: Glyph 2327 missing from current font.\n",
            "  font.set_text(s, 0.0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:214: RuntimeWarning: Glyph 2306 missing from current font.\n",
            "  font.set_text(s, 0.0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:214: RuntimeWarning: Glyph 2330 missing from current font.\n",
            "  font.set_text(s, 0.0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:214: RuntimeWarning: Glyph 2346 missing from current font.\n",
            "  font.set_text(s, 0.0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:214: RuntimeWarning: Glyph 2381 missing from current font.\n",
            "  font.set_text(s, 0.0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:214: RuntimeWarning: Glyph 2335 missing from current font.\n",
            "  font.set_text(s, 0.0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:214: RuntimeWarning: Glyph 2325 missing from current font.\n",
            "  font.set_text(s, 0.0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:214: RuntimeWarning: Glyph 2310 missing from current font.\n",
            "  font.set_text(s, 0.0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:214: RuntimeWarning: Glyph 2312 missing from current font.\n",
            "  font.set_text(s, 0.0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:214: RuntimeWarning: Glyph 2361 missing from current font.\n",
            "  font.set_text(s, 0.0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:214: RuntimeWarning: Glyph 2376 missing from current font.\n",
            "  font.set_text(s, 0.0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:214: RuntimeWarning: Glyph 2404 missing from current font.\n",
            "  font.set_text(s, 0.0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:214: RuntimeWarning: Glyph 2340 missing from current font.\n",
            "  font.set_text(s, 0.0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:214: RuntimeWarning: Glyph 2341 missing from current font.\n",
            "  font.set_text(s, 0.0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:214: RuntimeWarning: Glyph 2348 missing from current font.\n",
            "  font.set_text(s, 0.0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:214: RuntimeWarning: Glyph 2342 missing from current font.\n",
            "  font.set_text(s, 0.0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:214: RuntimeWarning: Glyph 2358 missing from current font.\n",
            "  font.set_text(s, 0.0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:214: RuntimeWarning: Glyph 2309 missing from current font.\n",
            "  font.set_text(s, 0.0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:214: RuntimeWarning: Glyph 2368 missing from current font.\n",
            "  font.set_text(s, 0.0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:214: RuntimeWarning: Glyph 2344 missing from current font.\n",
            "  font.set_text(s, 0.0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:214: RuntimeWarning: Glyph 2311 missing from current font.\n",
            "  font.set_text(s, 0.0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:214: RuntimeWarning: Glyph 2369 missing from current font.\n",
            "  font.set_text(s, 0.0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:214: RuntimeWarning: Glyph 2359 missing from current font.\n",
            "  font.set_text(s, 0.0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:214: RuntimeWarning: Glyph 2351 missing from current font.\n",
            "  font.set_text(s, 0.0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:214: RuntimeWarning: Glyph 2375 missing from current font.\n",
            "  font.set_text(s, 0.0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:214: RuntimeWarning: Glyph 2326 missing from current font.\n",
            "  font.set_text(s, 0.0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:214: RuntimeWarning: Glyph 2380 missing from current font.\n",
            "  font.set_text(s, 0.0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:214: RuntimeWarning: Glyph 2319 missing from current font.\n",
            "  font.set_text(s, 0.0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:214: RuntimeWarning: Glyph 2337 missing from current font.\n",
            "  font.set_text(s, 0.0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:214: RuntimeWarning: Glyph 2370 missing from current font.\n",
            "  font.set_text(s, 0.0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:214: RuntimeWarning: Glyph 2349 missing from current font.\n",
            "  font.set_text(s, 0.0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:214: RuntimeWarning: Glyph 2313 missing from current font.\n",
            "  font.set_text(s, 0.0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:214: RuntimeWarning: Glyph 2339 missing from current font.\n",
            "  font.set_text(s, 0.0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:214: RuntimeWarning: Glyph 2377 missing from current font.\n",
            "  font.set_text(s, 0.0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:214: RuntimeWarning: Glyph 2336 missing from current font.\n",
            "  font.set_text(s, 0.0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:214: RuntimeWarning: Glyph 2331 missing from current font.\n",
            "  font.set_text(s, 0.0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:214: RuntimeWarning: Glyph 2343 missing from current font.\n",
            "  font.set_text(s, 0.0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:214: RuntimeWarning: Glyph 2328 missing from current font.\n",
            "  font.set_text(s, 0.0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:214: RuntimeWarning: Glyph 2364 missing from current font.\n",
            "  font.set_text(s, 0.0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:214: RuntimeWarning: Glyph 2338 missing from current font.\n",
            "  font.set_text(s, 0.0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:214: RuntimeWarning: Glyph 2347 missing from current font.\n",
            "  font.set_text(s, 0.0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:214: RuntimeWarning: Glyph 2371 missing from current font.\n",
            "  font.set_text(s, 0.0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:214: RuntimeWarning: Glyph 2323 missing from current font.\n",
            "  font.set_text(s, 0.0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:214: RuntimeWarning: Glyph 2320 missing from current font.\n",
            "  font.set_text(s, 0.0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:214: RuntimeWarning: Glyph 2333 missing from current font.\n",
            "  font.set_text(s, 0.0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:183: RuntimeWarning: Glyph 2332 missing from current font.\n",
            "  font.set_text(s, 0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:183: RuntimeWarning: Glyph 2367 missing from current font.\n",
            "  font.set_text(s, 0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:183: RuntimeWarning: Glyph 2354 missing from current font.\n",
            "  font.set_text(s, 0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:183: RuntimeWarning: Glyph 2360 missing from current font.\n",
            "  font.set_text(s, 0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:183: RuntimeWarning: Glyph 2379 missing from current font.\n",
            "  font.set_text(s, 0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:183: RuntimeWarning: Glyph 2350 missing from current font.\n",
            "  font.set_text(s, 0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:183: RuntimeWarning: Glyph 2357 missing from current font.\n",
            "  font.set_text(s, 0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:183: RuntimeWarning: Glyph 2366 missing from current font.\n",
            "  font.set_text(s, 0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:183: RuntimeWarning: Glyph 2352 missing from current font.\n",
            "  font.set_text(s, 0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:183: RuntimeWarning: Glyph 2327 missing from current font.\n",
            "  font.set_text(s, 0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:183: RuntimeWarning: Glyph 2306 missing from current font.\n",
            "  font.set_text(s, 0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:183: RuntimeWarning: Glyph 2330 missing from current font.\n",
            "  font.set_text(s, 0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:183: RuntimeWarning: Glyph 2346 missing from current font.\n",
            "  font.set_text(s, 0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:183: RuntimeWarning: Glyph 2381 missing from current font.\n",
            "  font.set_text(s, 0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:183: RuntimeWarning: Glyph 2335 missing from current font.\n",
            "  font.set_text(s, 0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:183: RuntimeWarning: Glyph 2325 missing from current font.\n",
            "  font.set_text(s, 0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:183: RuntimeWarning: Glyph 2310 missing from current font.\n",
            "  font.set_text(s, 0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:183: RuntimeWarning: Glyph 2312 missing from current font.\n",
            "  font.set_text(s, 0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:183: RuntimeWarning: Glyph 2361 missing from current font.\n",
            "  font.set_text(s, 0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:183: RuntimeWarning: Glyph 2376 missing from current font.\n",
            "  font.set_text(s, 0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:183: RuntimeWarning: Glyph 2404 missing from current font.\n",
            "  font.set_text(s, 0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:183: RuntimeWarning: Glyph 2340 missing from current font.\n",
            "  font.set_text(s, 0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:183: RuntimeWarning: Glyph 2341 missing from current font.\n",
            "  font.set_text(s, 0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:183: RuntimeWarning: Glyph 2348 missing from current font.\n",
            "  font.set_text(s, 0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:183: RuntimeWarning: Glyph 2342 missing from current font.\n",
            "  font.set_text(s, 0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:183: RuntimeWarning: Glyph 2358 missing from current font.\n",
            "  font.set_text(s, 0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:183: RuntimeWarning: Glyph 2309 missing from current font.\n",
            "  font.set_text(s, 0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:183: RuntimeWarning: Glyph 2368 missing from current font.\n",
            "  font.set_text(s, 0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:183: RuntimeWarning: Glyph 2344 missing from current font.\n",
            "  font.set_text(s, 0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:183: RuntimeWarning: Glyph 2311 missing from current font.\n",
            "  font.set_text(s, 0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:183: RuntimeWarning: Glyph 2369 missing from current font.\n",
            "  font.set_text(s, 0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:183: RuntimeWarning: Glyph 2359 missing from current font.\n",
            "  font.set_text(s, 0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:183: RuntimeWarning: Glyph 2351 missing from current font.\n",
            "  font.set_text(s, 0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:183: RuntimeWarning: Glyph 2375 missing from current font.\n",
            "  font.set_text(s, 0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:183: RuntimeWarning: Glyph 2326 missing from current font.\n",
            "  font.set_text(s, 0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:183: RuntimeWarning: Glyph 2380 missing from current font.\n",
            "  font.set_text(s, 0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:183: RuntimeWarning: Glyph 2319 missing from current font.\n",
            "  font.set_text(s, 0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:183: RuntimeWarning: Glyph 2337 missing from current font.\n",
            "  font.set_text(s, 0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:183: RuntimeWarning: Glyph 2370 missing from current font.\n",
            "  font.set_text(s, 0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:183: RuntimeWarning: Glyph 2349 missing from current font.\n",
            "  font.set_text(s, 0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:183: RuntimeWarning: Glyph 2313 missing from current font.\n",
            "  font.set_text(s, 0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:183: RuntimeWarning: Glyph 2339 missing from current font.\n",
            "  font.set_text(s, 0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:183: RuntimeWarning: Glyph 2377 missing from current font.\n",
            "  font.set_text(s, 0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:183: RuntimeWarning: Glyph 2336 missing from current font.\n",
            "  font.set_text(s, 0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:183: RuntimeWarning: Glyph 2331 missing from current font.\n",
            "  font.set_text(s, 0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:183: RuntimeWarning: Glyph 2343 missing from current font.\n",
            "  font.set_text(s, 0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:183: RuntimeWarning: Glyph 2328 missing from current font.\n",
            "  font.set_text(s, 0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:183: RuntimeWarning: Glyph 2364 missing from current font.\n",
            "  font.set_text(s, 0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:183: RuntimeWarning: Glyph 2338 missing from current font.\n",
            "  font.set_text(s, 0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:183: RuntimeWarning: Glyph 2347 missing from current font.\n",
            "  font.set_text(s, 0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:183: RuntimeWarning: Glyph 2371 missing from current font.\n",
            "  font.set_text(s, 0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:183: RuntimeWarning: Glyph 2323 missing from current font.\n",
            "  font.set_text(s, 0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:183: RuntimeWarning: Glyph 2320 missing from current font.\n",
            "  font.set_text(s, 0, flags=flags)\n",
            "/usr/local/lib/python3.6/dist-packages/matplotlib/backends/backend_agg.py:183: RuntimeWarning: Glyph 2333 missing from current font.\n",
            "  font.set_text(s, 0, flags=flags)\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1152x1152 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "FcUBQGs3Ha7t",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "X_train = df['stopwords']\n",
        "Y_train = df['Category']"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "WFfOT0G3Z26o",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def preprocess1(x):\n",
        "  x = remove_punct(x)\n",
        "  x = generate(nltk.word_tokenize(x))\n",
        "  x = remove_stop_words(nltk.word_tokenize(x))\n",
        "  return x"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "fMtrjVrbULGy",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from sklearn.pipeline import Pipeline\n",
        "from sklearn import linear_model\n",
        "\n",
        "from sklearn.feature_extraction.text import TfidfVectorizer"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "j6wTDIPBURHQ",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "pipeline = Pipeline([\n",
        "    \n",
        "    ('tfidf',TfidfVectorizer(analyzer=\"word\",ngram_range=(1,3),max_features=5000 )),\n",
        "    ('clf', linear_model.LogisticRegression(C=1e5))\n",
        "])"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "8cDcCR5yUmLi",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 408
        },
        "outputId": "5a511e3c-1877-4229-a6cc-22bf308f4894"
      },
      "source": [
        "pipeline.fit(X_train, Y_train)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Pipeline(memory=None,\n",
              "         steps=[('tfidf',\n",
              "                 TfidfVectorizer(analyzer='word', binary=False,\n",
              "                                 decode_error='strict',\n",
              "                                 dtype=<class 'numpy.float64'>,\n",
              "                                 encoding='utf-8', input='content',\n",
              "                                 lowercase=True, max_df=1.0, max_features=5000,\n",
              "                                 min_df=1, ngram_range=(1, 3), norm='l2',\n",
              "                                 preprocessor=None, smooth_idf=True,\n",
              "                                 stop_words=None, strip_accents=None,\n",
              "                                 sublinear_tf=False,\n",
              "                                 token_pattern='(?u)\\\\b\\\\w\\\\w+\\\\b',\n",
              "                                 tokenizer=None, use_idf=True,\n",
              "                                 vocabulary=None)),\n",
              "                ('clf',\n",
              "                 LogisticRegression(C=100000.0, class_weight=None, dual=False,\n",
              "                                    fit_intercept=True, intercept_scaling=1,\n",
              "                                    l1_ratio=None, max_iter=100,\n",
              "                                    multi_class='auto', n_jobs=None,\n",
              "                                    penalty='l2', random_state=None,\n",
              "                                    solver='lbfgs', tol=0.0001, verbose=0,\n",
              "                                    warm_start=False))],\n",
              "         verbose=False)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 341
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "3OBzPQr4UxBE",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 292
        },
        "outputId": "fcd7f5a4-f9e6-428b-85cd-07ddd05537ca"
      },
      "source": [
        "!pip install googletrans"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Requirement already satisfied: googletrans in /usr/local/lib/python3.6/dist-packages (3.0.0)\n",
            "Requirement already satisfied: httpx==0.13.3 in /usr/local/lib/python3.6/dist-packages (from googletrans) (0.13.3)\n",
            "Requirement already satisfied: chardet==3.* in /usr/local/lib/python3.6/dist-packages (from httpx==0.13.3->googletrans) (3.0.4)\n",
            "Requirement already satisfied: sniffio in /usr/local/lib/python3.6/dist-packages (from httpx==0.13.3->googletrans) (1.1.0)\n",
            "Requirement already satisfied: certifi in /usr/local/lib/python3.6/dist-packages (from httpx==0.13.3->googletrans) (2020.6.20)\n",
            "Requirement already satisfied: rfc3986<2,>=1.3 in /usr/local/lib/python3.6/dist-packages (from httpx==0.13.3->googletrans) (1.4.0)\n",
            "Requirement already satisfied: hstspreload in /usr/local/lib/python3.6/dist-packages (from httpx==0.13.3->googletrans) (2020.7.17)\n",
            "Requirement already satisfied: httpcore==0.9.* in /usr/local/lib/python3.6/dist-packages (from httpx==0.13.3->googletrans) (0.9.1)\n",
            "Requirement already satisfied: idna==2.* in /usr/local/lib/python3.6/dist-packages (from httpx==0.13.3->googletrans) (2.10)\n",
            "Requirement already satisfied: contextvars>=2.1; python_version < \"3.7\" in /usr/local/lib/python3.6/dist-packages (from sniffio->httpx==0.13.3->googletrans) (2.4)\n",
            "Requirement already satisfied: h11<0.10,>=0.8 in /usr/local/lib/python3.6/dist-packages (from httpcore==0.9.*->httpx==0.13.3->googletrans) (0.9.0)\n",
            "Requirement already satisfied: h2==3.* in /usr/local/lib/python3.6/dist-packages (from httpcore==0.9.*->httpx==0.13.3->googletrans) (3.2.0)\n",
            "Requirement already satisfied: immutables>=0.9 in /usr/local/lib/python3.6/dist-packages (from contextvars>=2.1; python_version < \"3.7\"->sniffio->httpx==0.13.3->googletrans) (0.14)\n",
            "Requirement already satisfied: hyperframe<6,>=5.2.0 in /usr/local/lib/python3.6/dist-packages (from h2==3.*->httpcore==0.9.*->httpx==0.13.3->googletrans) (5.2.0)\n",
            "Requirement already satisfied: hpack<4,>=3.0 in /usr/local/lib/python3.6/dist-packages (from h2==3.*->httpcore==0.9.*->httpx==0.13.3->googletrans) (3.0.0)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "NrkafPjpW1W1",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "ae30fa3d-607c-4cfd-df01-33adfcf83e88"
      },
      "source": [
        "from googletrans import Translator\n",
        "translator = Translator()\n",
        "t = translator.translate('The number of people infected with this deadly virus has increased to 85,975 in the state after 3,007 new cases of coronavirus infection were reported in Maharashtra on Sunday. A health official said that after the death of 91 more people due to infection in the state, the death toll in the state has increased to 3,060. The official said that the cases of corona virus infection in Maharashtra have exceeded the cases of China. There are 83,036 confirmed cases of infection in China. 39,314 people have recovered in Maharashtra, 3,060 people have died and 43,591 peopleare undergoing treatment.',dest=\"hi\")\n",
        "pipeline.predict([t.text])"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([1])"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 343
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Dsi4ccg_Xi1I",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 54
        },
        "outputId": "20c5acab-2c5a-4773-c906-84fb054bc7d7"
      },
      "source": [
        "from googletrans import Translator\n",
        "translator = Translator()\n",
        "t = translator.translate(preprocess1('flynn hillary clinton big woman campus breitbart daniel j flynnever get feeling life circle roundabout rather head straight line toward intended destination hillary clinton remains big woman campus leafy liberal wellesley massachusetts everywhere else vote likely inauguration dress remainder day way miss havisham forever wore wedding dress speaking great expectations hillary rodham overflowed 48 year ago first addressed wellesley graduating class the president college informed gathered 1969 student needed debate far i could ascertain spokesman kind like democratic primary 2016 minus term unknown even seven sisters school i glad miss adams made clear i speaking today u 400 u miss rodham told classmate after appointing edger bergen charlie mccarthys mortimer snerds attendance bespectacled granny glass awarding matronly wisdom least john lennon wisdom took issue previous speaker despite becoming first win election seat u s senate since reconstruction edward brooke came criticism calling empathy goal protestors criticized tactic though clinton senior thesis saul alinsky lamented black power demagogue elitist arrogance repressive intolerance within new left similar word coming republican necessitated brief rebuttal trust rodham ironically observed 1969 one word i asked class rehearsal wanted say everyone came said talk trust talk lack trust u way feel others talk trust bust what say what say feeling permeates generation perhaps even understood distrusted the trust bust certainly busted clintons 2016 plan she certainly even understand people distrusted after whitewater travelgate vast conspiracy benghazi missing email clinton found distrusted voice friday there load compromising road broadening political horizon and distrust american people trump edged 48 percent 38 percent question immediately prior novembers election stood major reason closing horizon clinton described vanquisher supporter embracing lie con alternative fact assault truth reason she failed explain american people chose lie truth as history major among today know well people power invent fact attack question mark beginning end free society offered that hyperbole like many people emerge 1960s hillary clinton embarked upon long strange trip from high school goldwater girl wellesley college republican president democratic politician clinton drank time place gave degree more significantly went idealist cynic comparison two wellesley commencement address show way back lamented long leader viewed politics art possible challenge practice politics art making appears impossible possible now big woman campus odd woman white house wonder current station even possible why arent i 50 point ahead asked september in may asks isnt president the woman famously dubbed congenital liar bill safire concludes lie mind getting stood election day like finding jilted bride wedding day inspires dangerous delusion'),dest=\"hi\")\n",
        "pipeline.predict([preprocess(t.text)])\n",
        "print(t.text)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "उड़ता हिलेरी क्लिंटन बड़ी महिला परिसर breitbart daniel j flynnever को जीवन चक्र गोल चक्कर के रूप में महसूस करना पड़ता है, लेकिन इसका मतलब है कि गंतव्य के लिए सीधी रेखा, हिलेरी क्लिंटन बड़ी महिला परिसर पत्तेदार लिबरल वेस्ली मासैचियन हर जगह रहती है, जहाँ से उद्घाटन की संभावना शेष रहती है जिस तरह से मिस हैविशम हमेशा के लिए शादी की पोशाक पहनकर बड़ी उम्मीदें रखती है। हिलेरी रोथम 48 साल पहले बह निकला था, जिसने पहले से ही स्नातक की उपाधि प्राप्त कर ली थी। राष्ट्रपति कॉलेज ने 1969 के छात्रों को बहस की जरूरत बताई थी कि मुझे लोकतांत्रिक प्राथमिक 2016 की तरह ही प्रवक्ता की तरह पता लगाया जा सकता है कि अज्ञात सात बहनें भी स्कूली हैं, मुझे खुशी है कि एडम्स स्पष्ट हैं मैं आज ही बोल रहा हूं। यू 400 यू मिस रॉडहम ने एडगर बर्गेन चार्ली मकार्थिस मोर्टिमर सिनर्ड्स की उपस्थिति के बाद सहपाठी से कहा कि उपस्थिति के बारे में बताए गए नानी ग्लास को मैट्रनली ज्ञान प्रदान करते हुए कम से कम जॉन लेनन ज्ञान ने पिछले स्पीकर को पहली बार चुनाव सीट देने के बावजूद हमें स्पीकर बना दिया क्योंकि पुनर्निर्माण एडवर्ड ब्रुक की आलोचना के बाद से सीनेट की आलोचना हुई। एमपैथी गोल प्रदर्शनकारियों ने रणनीति की आलोचना की, हालांकि क्लिंटन वरिष्ठ थीसिस सूल अलिन्स्की ने ब्लैक पॉवर डेमोगॉग एलीटिस्ट एगरिस्ट दमनकारी असहिष्णुता को नए बाएँ शब्दों के भीतर रखा है। इसी तरह के शब्द रिपब्लिकन आवश्यक संक्षिप्त खंडन्यास ट्रस्ट रॉडहैम ने 1969 में देखा एक शब्द मैंने कक्षा रिहर्सल चाहा था, जिसमें कहा गया था कि हर कोई विश्वास की कमी की बात करता है। यू वे महसूस करते हैं कि अन्य लोग ट्रस्ट बस्ट बोलते हैं जो कहते हैं कि क्या महसूस करती है कि परमिट जनरेशन जनरेशन शायद अविश्वास को भी समझ गई है बस्ट निश्चित रूप से पर्दाफाश क्लिंटन 2016 की योजना वह निश्चित रूप से लोगों को भी समझती है जो व्हाइटवॉटर ट्रैवलगेट विशाल साजिश के बाद अविश्वासित होते हैं बेन्थाजी लापता ईमेल क्लिंट पाया डिस्ट्रेस्ड वॉयस फ्राइडे लोड लोडिंग रोड चौड़ीकरण राजनीतिक क्षितिज और अविश्वास अमेरिकी लोगों ने ट्रम्प को 48 प्रतिशत 38 प्रतिशत प्रश्न के तुरंत पहले नुक्कड़ सभा चुनाव के लिए खड़ा कर दिया, प्रमुख कारण क्षितिज क्लिंटन को बंद करना बताया गया था, जिसमें वेक्किर समर्थक झूठ बोल रहे थे। एक लोगों ने झूठ को सच के रूप में चुना है, जो आज के इतिहास के प्रमुख के रूप में अच्छी तरह से जानते हैं कि लोग पावर इनवेशन फैक्ट अटैक क्वेश्चन अटैक एंड मार्क फ्री सोसाइटी की शुरुआत करते हैं, जिसमें कहा गया है कि हाईबोल जैसे कई लोग 1960 के दशक में हिलेरी क्लिंटन के रूप में उभरे हैं, जो हाई स्कूल गोल्डवॉटर गर्ल वेलस्ले कॉलेज रिपब्लिकन राष्ट्रपति लोकतांत्रिक राजनीतिज्ञ क्लिंटन से लंबी अजीब यात्रा करते हैं। प्याऊ का समय दिया गया, डिग्री ने काफी हद तक आदर्शवादी निंदक तुलना दो वेलसली स्टार्टिंग एड्रेस शो को वापस ले लिया, जिस पर लंबे समय तक चलने वाला नेता राजनीति कला संभव चुनौती अभ्यास राजनीति कला बनाना असंभव प्रतीत होता है अब बड़ी महिला कैंपस विषम महिला व्हाइट हाउस आश्चर्य वर्तमान स्टेशन यहां तक ​​कि संभव नहीं है - arent i 50 आगे से पूछे जाने वाले सिपाही से पूछते हैं कि इस्लाम राष्ट्रपति ने महिला को प्रसिद्ध जन्मजात झूठे बिल को सुरक्षित रखने के लिए झूठे मन से चुनावी दिन की तरह चुना हुआ है, जैसे कि झुकी हुई दुल्हन की शादी का दिन खतरनाक भ्रम को प्रेरित करता है\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "RFUwPUd9YEmu",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "e4f77447-4805-4640-d7ed-1e3ac6bc9027"
      },
      "source": [
        "from googletrans import Translator\n",
        "translator = Translator()\n",
        "t = translator.translate('The vaccine for corona has already come',dest=\"hi\")\n",
        "pipeline.predict([preprocess1(t.text)])"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([0])"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 345
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "8NZX_93JYS2C",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "8ac0a889-39cf-45b5-971c-e2066e343da7"
      },
      "source": [
        "from googletrans import Translator\n",
        "translator = Translator()\n",
        "t = translator.translate('eating vegetarian food and eliminating meat from your diet could prevent you getting coronavirus.Indian cricketer Sachin Tendulkar had asked for meat shops to be shut.',dest=\"hi\")\n",
        "pipeline.predict([preprocess1(t.text)])"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([0])"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 347
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "iC59LdFGfOU6",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 120
        },
        "outputId": "d44608fa-f004-4947-81e0-3c1140528903"
      },
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/content/drive')"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Go to this URL in a browser: https://accounts.google.com/o/oauth2/auth?client_id=947318989803-6bn6qk8qdgf4n4g3pfee6491hc0brc4i.apps.googleusercontent.com&redirect_uri=urn%3aietf%3awg%3aoauth%3a2.0%3aoob&response_type=code&scope=email%20https%3a%2f%2fwww.googleapis.com%2fauth%2fdocs.test%20https%3a%2f%2fwww.googleapis.com%2fauth%2fdrive%20https%3a%2f%2fwww.googleapis.com%2fauth%2fdrive.photos.readonly%20https%3a%2f%2fwww.googleapis.com%2fauth%2fpeopleapi.readonly\n",
            "\n",
            "Enter your authorization code:\n",
            "··········\n",
            "Mounted at /content/drive\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "yFVgDmtVfL0y",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import numpy \n",
        "\n",
        "X = numpy.array(list(df['stopwords']))"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_mmyX9-rgi_z",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 33
        },
        "outputId": "3ffdd114-7028-4bc1-eebf-d7b5bf451315"
      },
      "source": [
        "X.shape"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(843,)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 21
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "I_7kpShkgkB-",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "y = numpy.array(list(df['Category']))"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "rdrfAMKfgn68",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 33
        },
        "outputId": "6c5d924d-3999-44b9-b62f-087ae98c02d3"
      },
      "source": [
        "y.shape"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(843,)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 23
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "D4hX0D-mgoui",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from sklearn.model_selection import train_test_split\n",
        "df_train, df_test, Ytrain, Ytest = train_test_split(X, y, test_size=0.33)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "6vW8eeQLgtND",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from tensorflow.keras.preprocessing.text import Tokenizer\n",
        "from tensorflow.keras.preprocessing.sequence import pad_sequences\n",
        "from tensorflow.keras.layers import Dense, Input, GlobalMaxPooling1D, SpatialDropout1D, Conv1D, MaxPooling1D\n",
        "from tensorflow.keras.layers import LSTM, Embedding\n",
        "from tensorflow.keras.models import Model\n",
        "MAX_VOCAB_SIZE = 20000\n",
        "tokenizer = Tokenizer(num_words=MAX_VOCAB_SIZE)\n",
        "tokenizer.fit_on_texts(df_train)\n",
        "sequences_train = tokenizer.texts_to_sequences(df_train)\n",
        "sequences_test = tokenizer.texts_to_sequences(df_test)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Y0jHmVf9gwHm",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 33
        },
        "outputId": "0cd699ce-8f83-4688-98f8-1bc2c5df2ce0"
      },
      "source": [
        "len(tokenizer.word_index)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "9482"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 26
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "It8xY3WBgzYX",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 33
        },
        "outputId": "fc0dd2d4-437a-4f8d-d25d-7b87f1c1f2fb"
      },
      "source": [
        "data_train = pad_sequences(sequences_train)\n",
        "print('Shape of data train tensor:', data_train.shape)\n",
        "\n",
        "# get sequence length\n",
        "T = data_train.shape[1]"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Shape of data train tensor: (564, 1581)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "kX4lf2JZg3ky",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 33
        },
        "outputId": "3cc67902-fe9f-491e-cfc0-b59dfaf19b9b"
      },
      "source": [
        "data_test = pad_sequences(sequences_test, maxlen=T)\n",
        "print('Shape of data test tensor:', data_test.shape)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Shape of data test tensor: (279, 1581)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "qu1UhvkVg8LT",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 133
        },
        "outputId": "a6cc6310-154b-435c-98ea-b3cd1a399848"
      },
      "source": [
        "data_test"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[   0,    0,    0, ..., 2779, 1621,    1],\n",
              "       [   0,    0,    0, ..., 1145,  603,  410],\n",
              "       [   0,    0,    0, ...,  189,  224,  225],\n",
              "       ...,\n",
              "       [   0,    0,    0, ...,    7,    6,    1],\n",
              "       [   0,    0,    0, ...,  137,   78,  424],\n",
              "       [   0,    0,    0, ...,  189,  224,  225]], dtype=int32)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 29
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "6Pnm4iwmg-_C",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 863
        },
        "outputId": "3ce96343-8bf7-4ef3-d0b4-f2c76340e041"
      },
      "source": [
        "from tensorflow.keras.callbacks import EarlyStopping\n",
        "from tensorflow.keras.optimizers import Adam\n",
        "\n",
        "early_stop = EarlyStopping(monitor='val_loss',patience=2)\n",
        "# Create the model\n",
        "\n",
        "V= len(tokenizer.word_index)\n",
        "# We get to choose embedding dimensionality\n",
        "D = 20\n",
        "\n",
        "# Hidden state dimensionality\n",
        "M = 100\n",
        "\n",
        "# Note: we actually want to the size of the embedding to (V + 1) x D,\n",
        "# because the first index starts from 1 and not 0.\n",
        "# Thus, if the final index of the embedding matrix is V,\n",
        "# then it actually must have size V + 1.\n",
        "\n",
        "i = Input(shape=(T,))\n",
        "x = Embedding(V + 1, D)(i)\n",
        "x = SpatialDropout1D(0.2)(x)\n",
        "x = LSTM(M, return_sequences=True)(x)\n",
        "x = Dense(40,activation='relu')(x)\n",
        "x = GlobalMaxPooling1D()(x)\n",
        "x = Dense(1, activation='sigmoid')(x)\n",
        "\n",
        "\n",
        "model = Model(i, x)\n",
        "# Compile and fit\n",
        "model.compile(\n",
        "  loss='binary_crossentropy',\n",
        "  optimizer='adam',\n",
        "  metrics=['accuracy']\n",
        ")\n",
        "\n",
        "\n",
        "print('Training model...')\n",
        "r = model.fit(\n",
        "  data_train,\n",
        "  Ytrain,\n",
        "  epochs=20,\n",
        "  validation_data=(data_test, Ytest),\n",
        "  callbacks=[early_stop]\n",
        "  #validation_split=0.1\n",
        ")\n",
        "# Plot loss per iteration\n",
        "import matplotlib.pyplot as plt\n",
        "plt.plot(r.history['loss'], label='loss')\n",
        "plt.plot(r.history['val_loss'], label='val_loss')\n",
        "plt.legend()\n",
        "plt.savefig('LSTMloss.png', dpi=300, bbox_inches='tight')\n",
        "plt.show()\n",
        "plt.plot(r.history['accuracy'], label='acc')\n",
        "plt.plot(r.history['val_accuracy'], label='val_acc')\n",
        "plt.legend()\n",
        "plt.savefig('LSTMAccu.png', dpi=300, bbox_inches='tight')\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Training model...\n",
            "Epoch 1/20\n",
            "18/18 [==============================] - 2s 96ms/step - loss: 0.6820 - accuracy: 0.5585 - val_loss: 0.5726 - val_accuracy: 0.9211\n",
            "Epoch 2/20\n",
            "18/18 [==============================] - 1s 72ms/step - loss: 0.6402 - accuracy: 0.7571 - val_loss: 0.6098 - val_accuracy: 0.7599\n",
            "Epoch 3/20\n",
            "18/18 [==============================] - 1s 71ms/step - loss: 0.4886 - accuracy: 0.8688 - val_loss: 0.2779 - val_accuracy: 0.9391\n",
            "Epoch 4/20\n",
            "18/18 [==============================] - 1s 71ms/step - loss: 0.3649 - accuracy: 0.8688 - val_loss: 0.4863 - val_accuracy: 0.8638\n",
            "Epoch 5/20\n",
            "18/18 [==============================] - 1s 71ms/step - loss: 0.2735 - accuracy: 0.9521 - val_loss: 0.1818 - val_accuracy: 0.9785\n",
            "Epoch 6/20\n",
            "18/18 [==============================] - 1s 72ms/step - loss: 0.1274 - accuracy: 0.9681 - val_loss: 0.0943 - val_accuracy: 0.9785\n",
            "Epoch 7/20\n",
            "18/18 [==============================] - 1s 74ms/step - loss: 0.0591 - accuracy: 0.9840 - val_loss: 0.0818 - val_accuracy: 0.9928\n",
            "Epoch 8/20\n",
            "18/18 [==============================] - 1s 72ms/step - loss: 0.0280 - accuracy: 0.9965 - val_loss: 0.0671 - val_accuracy: 0.9928\n",
            "Epoch 9/20\n",
            "18/18 [==============================] - 1s 72ms/step - loss: 0.0224 - accuracy: 0.9965 - val_loss: 0.0683 - val_accuracy: 0.9892\n",
            "Epoch 10/20\n",
            "18/18 [==============================] - 1s 71ms/step - loss: 0.0207 - accuracy: 0.9965 - val_loss: 0.0715 - val_accuracy: 0.9892\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "NSe7k4B0hJHX",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 796
        },
        "outputId": "c4afee74-62a3-4a2e-f5ab-2e4138d42215"
      },
      "source": [
        "from tensorflow.keras.callbacks import EarlyStopping\n",
        "from tensorflow.keras.optimizers import Adam\n",
        "\n",
        "early_stop = EarlyStopping(monitor='accuracy',patience=2)\n",
        "i = Input(shape=(T,))\n",
        "x = Embedding(V + 1, D)(i)\n",
        "x = Conv1D(32, 3, activation='relu')(x)\n",
        "x = MaxPooling1D(3)(x)\n",
        "x = Conv1D(64, 3, activation='relu')(x)\n",
        "x = MaxPooling1D(3)(x)\n",
        "x = Conv1D(128, 3, activation='relu')(x)\n",
        "x = GlobalMaxPooling1D()(x)\n",
        "x = Dense(1, activation='sigmoid')(x)\n",
        "\n",
        "model = Model(i, x)\n",
        "\n",
        "model.compile(Adam(lr=0.001),\n",
        "              loss='binary_crossentropy',\n",
        "              metrics=['accuracy'])\n",
        "\n",
        "print('Training model...')\n",
        "r = model.fit(\n",
        "  data_train,\n",
        "  Ytrain,\n",
        "  epochs=10,\n",
        "  validation_data=(data_test, Ytest), callbacks=[early_stop]\n",
        ")\n",
        "import matplotlib.pyplot as plt\n",
        "plt.plot(r.history['loss'], label='loss')\n",
        "plt.plot(r.history['val_loss'], label='val_loss')\n",
        "plt.legend()\n",
        "plt.savefig('CNNloss.png', dpi=300, bbox_inches='tight')\n",
        "plt.show()\n",
        "plt.plot(r.history['accuracy'], label='acc')\n",
        "plt.plot(r.history['val_accuracy'], label='val_acc')\n",
        "plt.legend()\n",
        "plt.savefig('CNNaccu.png', dpi=300, bbox_inches='tight')\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Training model...\n",
            "Epoch 1/10\n",
            "18/18 [==============================] - 0s 18ms/step - loss: 0.6882 - accuracy: 0.4876 - val_loss: 0.6775 - val_accuracy: 0.8530\n",
            "Epoch 2/10\n",
            "18/18 [==============================] - 0s 11ms/step - loss: 0.6542 - accuracy: 0.7748 - val_loss: 0.6048 - val_accuracy: 0.7778\n",
            "Epoch 3/10\n",
            "18/18 [==============================] - 0s 12ms/step - loss: 0.4763 - accuracy: 0.8582 - val_loss: 0.2863 - val_accuracy: 0.9355\n",
            "Epoch 4/10\n",
            "18/18 [==============================] - 0s 12ms/step - loss: 0.1775 - accuracy: 0.9539 - val_loss: 0.1179 - val_accuracy: 0.9677\n",
            "Epoch 5/10\n",
            "18/18 [==============================] - 0s 12ms/step - loss: 0.0486 - accuracy: 0.9876 - val_loss: 0.0638 - val_accuracy: 0.9785\n",
            "Epoch 6/10\n",
            "18/18 [==============================] - 0s 11ms/step - loss: 0.0206 - accuracy: 0.9965 - val_loss: 0.0622 - val_accuracy: 0.9785\n",
            "Epoch 7/10\n",
            "18/18 [==============================] - 0s 12ms/step - loss: 0.0199 - accuracy: 0.9965 - val_loss: 0.0583 - val_accuracy: 0.9785\n",
            "Epoch 8/10\n",
            "18/18 [==============================] - 0s 12ms/step - loss: 0.0193 - accuracy: 0.9965 - val_loss: 0.0577 - val_accuracy: 0.9785\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "JvA2FPg7hWoL",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 606
        },
        "outputId": "b0f6cb9a-979d-409b-d261-52db232132e5"
      },
      "source": [
        "from wordcloud import WordCloud, STOPWORDS \n",
        "cop=' '.join(df['stopwords'])\n",
        "\n",
        "wordcloud = WordCloud(width = 800, height = 800, \n",
        "                background_color ='white', \n",
        "                 \n",
        "                min_font_size = 10).generate(cop) \n",
        "  \n",
        "# plot the WordCloud image                        \n",
        "plt.figure(figsize = (8, 8), facecolor = None) \n",
        "plt.imshow(wordcloud) \n",
        "plt.axis(\"off\") \n",
        "plt.tight_layout(pad = 0) \n",
        "  \n",
        "plt.show() "
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 576x576 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Ugn4m606iCri",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    }
  ]
}