machine_learning_approach.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "kernelspec": {
      "display_name": "Python 3",
      "language": "python",
      "name": "python3"
    },
    "language_info": {
      "codemirror_mode": {
        "name": "ipython",
        "version": 3
      },
      "file_extension": ".py",
      "mimetype": "text/x-python",
      "name": "python",
      "nbconvert_exporter": "python",
      "pygments_lexer": "ipython3",
      "version": "3.7.9"
    },
    "papermill": {
      "default_parameters": {},
      "duration": 565.856184,
      "end_time": "2021-03-01T19:17:05.796650",
      "environment_variables": {},
      "exception": null,
      "input_path": "__notebook__.ipynb",
      "output_path": "__notebook__.ipynb",
      "parameters": {},
      "start_time": "2021-03-01T19:07:39.940466",
      "version": "2.2.2"
    },
    "colab": {
      "name": "machine-learning-approach.ipynb",
      "provenance": []
    }
  },
  "cells": [
    {
      "cell_type": "code",
      "metadata": {
        "_cell_guid": "b1076dfc-b9ad-4769-8c92-a6c4dae69d19",
        "_uuid": "8f2839f25d086af736a60e9eeb907d3b93b6e0e5",
        "execution": {
          "iopub.execute_input": "2021-03-01T19:07:44.538888Z",
          "iopub.status.busy": "2021-03-01T19:07:44.538135Z",
          "iopub.status.idle": "2021-03-01T19:07:45.964361Z",
          "shell.execute_reply": "2021-03-01T19:07:45.963212Z"
        },
        "papermill": {
          "duration": 1.45082,
          "end_time": "2021-03-01T19:07:45.964579",
          "exception": false,
          "start_time": "2021-03-01T19:07:44.513759",
          "status": "completed"
        },
        "tags": [],
        "id": "laX85Ie_VFcW"
      },
      "source": [
        "\n",
        "%matplotlib inline\n",
        "import re\n",
        "import matplotlib\n",
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "import pandas as pd\n",
        "from sklearn.model_selection import train_test_split\n",
        "from sklearn.feature_extraction.text import TfidfVectorizer\n",
        "from sklearn.naive_bayes import MultinomialNB\n",
        "from sklearn.metrics import accuracy_score\n",
        "from sklearn.multiclass import OneVsRestClassifier\n",
        "from nltk.corpus import stopwords\n",
        "stop_words = set(stopwords.words('english'))\n",
        "from sklearn.svm import LinearSVC\n",
        "from sklearn.linear_model import LogisticRegression\n",
        "from sklearn.pipeline import Pipeline\n",
        "import seaborn as sns"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-03-01T19:07:46.010283Z",
          "iopub.status.busy": "2021-03-01T19:07:46.009751Z",
          "iopub.status.idle": "2021-03-01T19:07:47.678565Z",
          "shell.execute_reply": "2021-03-01T19:07:47.677929Z"
        },
        "papermill": {
          "duration": 1.700144,
          "end_time": "2021-03-01T19:07:47.678703",
          "exception": false,
          "start_time": "2021-03-01T19:07:45.978559",
          "status": "completed"
        },
        "tags": [],
        "id": "52YArlrGVFcd"
      },
      "source": [
        "df = pd.read_csv(\"../input/nlp-coursework/1613373921-5e748a2d5fc288e9f69c5f86.csv\")"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-03-01T19:07:47.715243Z",
          "iopub.status.busy": "2021-03-01T19:07:47.713646Z",
          "iopub.status.idle": "2021-03-01T19:07:47.715858Z",
          "shell.execute_reply": "2021-03-01T19:07:47.716242Z"
        },
        "papermill": {
          "duration": 0.021268,
          "end_time": "2021-03-01T19:07:47.716370",
          "exception": false,
          "start_time": "2021-03-01T19:07:47.695102",
          "status": "completed"
        },
        "tags": [],
        "id": "UdBKizbCVFcd"
      },
      "source": [
        "label_cols = ['toxic', 'severe_toxic', 'obscene', 'threat', 'insult', 'identity_hate']"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-03-01T19:07:47.759409Z",
          "iopub.status.busy": "2021-03-01T19:07:47.758497Z",
          "iopub.status.idle": "2021-03-01T19:07:47.837511Z",
          "shell.execute_reply": "2021-03-01T19:07:47.836987Z"
        },
        "papermill": {
          "duration": 0.107119,
          "end_time": "2021-03-01T19:07:47.837661",
          "exception": false,
          "start_time": "2021-03-01T19:07:47.730542",
          "status": "completed"
        },
        "tags": [],
        "id": "916i92ABVFce",
        "outputId": "f501d392-2484-4dc8-e227-52d3268d2187"
      },
      "source": [
        "df['none'] = 1-df[label_cols].max(axis=1)\n",
        "df.describe()"
      ],
      "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>toxic</th>\n",
              "      <th>severe_toxic</th>\n",
              "      <th>obscene</th>\n",
              "      <th>threat</th>\n",
              "      <th>insult</th>\n",
              "      <th>identity_hate</th>\n",
              "      <th>none</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>count</th>\n",
              "      <td>159571.000000</td>\n",
              "      <td>159571.000000</td>\n",
              "      <td>159571.000000</td>\n",
              "      <td>159571.000000</td>\n",
              "      <td>159571.000000</td>\n",
              "      <td>159571.000000</td>\n",
              "      <td>159571.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>mean</th>\n",
              "      <td>0.095844</td>\n",
              "      <td>0.009996</td>\n",
              "      <td>0.052948</td>\n",
              "      <td>0.002996</td>\n",
              "      <td>0.049364</td>\n",
              "      <td>0.008805</td>\n",
              "      <td>0.898321</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>std</th>\n",
              "      <td>0.294379</td>\n",
              "      <td>0.099477</td>\n",
              "      <td>0.223931</td>\n",
              "      <td>0.054650</td>\n",
              "      <td>0.216627</td>\n",
              "      <td>0.093420</td>\n",
              "      <td>0.302226</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>min</th>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>25%</th>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>1.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>50%</th>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>1.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>75%</th>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>1.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>max</th>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "               toxic   severe_toxic        obscene         threat  \\\n",
              "count  159571.000000  159571.000000  159571.000000  159571.000000   \n",
              "mean        0.095844       0.009996       0.052948       0.002996   \n",
              "std         0.294379       0.099477       0.223931       0.054650   \n",
              "min         0.000000       0.000000       0.000000       0.000000   \n",
              "25%         0.000000       0.000000       0.000000       0.000000   \n",
              "50%         0.000000       0.000000       0.000000       0.000000   \n",
              "75%         0.000000       0.000000       0.000000       0.000000   \n",
              "max         1.000000       1.000000       1.000000       1.000000   \n",
              "\n",
              "              insult  identity_hate           none  \n",
              "count  159571.000000  159571.000000  159571.000000  \n",
              "mean        0.049364       0.008805       0.898321  \n",
              "std         0.216627       0.093420       0.302226  \n",
              "min         0.000000       0.000000       0.000000  \n",
              "25%         0.000000       0.000000       1.000000  \n",
              "50%         0.000000       0.000000       1.000000  \n",
              "75%         0.000000       0.000000       1.000000  \n",
              "max         1.000000       1.000000       1.000000  "
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 4
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-03-01T19:07:47.878914Z",
          "iopub.status.busy": "2021-03-01T19:07:47.878007Z",
          "iopub.status.idle": "2021-03-01T19:07:47.890697Z",
          "shell.execute_reply": "2021-03-01T19:07:47.891102Z"
        },
        "papermill": {
          "duration": 0.036828,
          "end_time": "2021-03-01T19:07:47.891252",
          "exception": false,
          "start_time": "2021-03-01T19:07:47.854424",
          "status": "completed"
        },
        "tags": [],
        "id": "fEnc3ZGDVFcf",
        "outputId": "3472b073-a854-47cd-d6ff-f0bfa9757e5a"
      },
      "source": [
        "df_toxic = df.drop(['id', 'comment_text'], axis=1)\n",
        "counts = []\n",
        "categories = list(df_toxic.columns.values)\n",
        "for i in categories:\n",
        "    counts.append((i, df_toxic[i].sum()))\n",
        "df_stats = pd.DataFrame(counts, columns=['category', 'number_of_comments'])\n",
        "df_stats"
      ],
      "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>category</th>\n",
              "      <th>number_of_comments</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>toxic</td>\n",
              "      <td>15294</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>severe_toxic</td>\n",
              "      <td>1595</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>obscene</td>\n",
              "      <td>8449</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>threat</td>\n",
              "      <td>478</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>insult</td>\n",
              "      <td>7877</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>identity_hate</td>\n",
              "      <td>1405</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6</th>\n",
              "      <td>none</td>\n",
              "      <td>143346</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "        category  number_of_comments\n",
              "0          toxic               15294\n",
              "1   severe_toxic                1595\n",
              "2        obscene                8449\n",
              "3         threat                 478\n",
              "4         insult                7877\n",
              "5  identity_hate                1405\n",
              "6           none              143346"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 5
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-03-01T19:07:47.927623Z",
          "iopub.status.busy": "2021-03-01T19:07:47.926853Z",
          "iopub.status.idle": "2021-03-01T19:07:48.119165Z",
          "shell.execute_reply": "2021-03-01T19:07:48.119600Z"
        },
        "papermill": {
          "duration": 0.212983,
          "end_time": "2021-03-01T19:07:48.119755",
          "exception": false,
          "start_time": "2021-03-01T19:07:47.906772",
          "status": "completed"
        },
        "tags": [],
        "id": "-rZWt1kuVFcf",
        "outputId": "fa428235-5991-4bbe-90fe-c9754dd6dc19"
      },
      "source": [
        "df_stats.plot(x='category', y='number_of_comments', kind='bar', legend=False, grid=True, figsize=(8, 5))\n",
        "plt.title(\"Number of comments per category\")\n",
        "plt.ylabel('# of Occurrences', fontsize=12)\n",
        "plt.xlabel('category', fontsize=12)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Text(0.5, 0, 'category')"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 6
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 576x360 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-03-01T19:07:48.163391Z",
          "iopub.status.busy": "2021-03-01T19:07:48.162593Z",
          "iopub.status.idle": "2021-03-01T19:07:48.346494Z",
          "shell.execute_reply": "2021-03-01T19:07:48.345984Z"
        },
        "papermill": {
          "duration": 0.209766,
          "end_time": "2021-03-01T19:07:48.346639",
          "exception": false,
          "start_time": "2021-03-01T19:07:48.136873",
          "status": "completed"
        },
        "tags": [],
        "id": "8EKNcFyGVFcg",
        "outputId": "0cd47f92-ab75-4345-d070-620cef4e6c5f"
      },
      "source": [
        "rowsums = df.iloc[:,2:].sum(axis=1)\n",
        "x=rowsums.value_counts()\n",
        "\n",
        "#plot\n",
        "plt.figure(figsize=(8,5))\n",
        "ax = sns.barplot(x.index, x.values)\n",
        "plt.title(\"Multiple categories per comment\")\n",
        "plt.ylabel('# of Occurrences', fontsize=12)\n",
        "plt.xlabel('# of categories', fontsize=12)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/opt/conda/lib/python3.7/site-packages/seaborn/_decorators.py:43: FutureWarning: Pass the following variables as keyword args: x, y. From version 0.12, the only valid positional argument will be `data`, and passing other arguments without an explicit keyword will result in an error or misinterpretation.\n",
            "  FutureWarning\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Text(0.5, 0, '# of categories')"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 7
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 576x360 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-03-01T19:07:48.437729Z",
          "iopub.status.busy": "2021-03-01T19:07:48.410888Z",
          "iopub.status.idle": "2021-03-01T19:07:48.789015Z",
          "shell.execute_reply": "2021-03-01T19:07:48.789470Z"
        },
        "papermill": {
          "duration": 0.423698,
          "end_time": "2021-03-01T19:07:48.789634",
          "exception": false,
          "start_time": "2021-03-01T19:07:48.365936",
          "status": "completed"
        },
        "tags": [],
        "id": "M2SSWsMpVFcg",
        "outputId": "0b72a1b1-830f-4f7f-b515-64660c80e294"
      },
      "source": [
        "lens = df.comment_text.str.len()\n",
        "lens.hist(bins = np.arange(0,5000,50))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<AxesSubplot:>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 8
        },
        {
          "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": {
        "execution": {
          "iopub.execute_input": "2021-03-01T19:07:48.836387Z",
          "iopub.status.busy": "2021-03-01T19:07:48.835565Z",
          "iopub.status.idle": "2021-03-01T19:07:48.869494Z",
          "shell.execute_reply": "2021-03-01T19:07:48.869930Z"
        },
        "papermill": {
          "duration": 0.061355,
          "end_time": "2021-03-01T19:07:48.870081",
          "exception": false,
          "start_time": "2021-03-01T19:07:48.808726",
          "status": "completed"
        },
        "tags": [],
        "id": "XFC8-ehPVFch",
        "outputId": "3ba3b24d-6219-45ac-cacb-e1f9f4a2f257"
      },
      "source": [
        "print('Number of missing comments in comment text:')\n",
        "df['comment_text'].isnull().sum()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Number of missing comments in comment text:\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 9
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-03-01T19:07:48.914320Z",
          "iopub.status.busy": "2021-03-01T19:07:48.913646Z",
          "iopub.status.idle": "2021-03-01T19:07:48.917013Z",
          "shell.execute_reply": "2021-03-01T19:07:48.917407Z"
        },
        "papermill": {
          "duration": 0.028348,
          "end_time": "2021-03-01T19:07:48.917561",
          "exception": false,
          "start_time": "2021-03-01T19:07:48.889213",
          "status": "completed"
        },
        "tags": [],
        "id": "xLUrGoIlVFch",
        "outputId": "9ce4a2ce-1774-4547-c59d-36d837427785"
      },
      "source": [
        "df['comment_text'][0]"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "\"Explanation\\nWhy the edits made under my username Hardcore Metallica Fan were reverted? They weren't vandalisms, just closure on some GAs after I voted at New York Dolls FAC. And please don't remove the template from the talk page since I'm retired now.89.205.38.27\""
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 10
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-03-01T19:07:48.960482Z",
          "iopub.status.busy": "2021-03-01T19:07:48.959919Z",
          "iopub.status.idle": "2021-03-01T19:07:48.964096Z",
          "shell.execute_reply": "2021-03-01T19:07:48.963658Z"
        },
        "papermill": {
          "duration": 0.02736,
          "end_time": "2021-03-01T19:07:48.964230",
          "exception": false,
          "start_time": "2021-03-01T19:07:48.936870",
          "status": "completed"
        },
        "tags": [],
        "id": "JonSxeCkVFch"
      },
      "source": [
        "categories = ['toxic', 'severe_toxic', 'obscene', 'threat', 'insult', 'identity_hate','none']"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-03-01T19:07:49.012170Z",
          "iopub.status.busy": "2021-03-01T19:07:49.010286Z",
          "iopub.status.idle": "2021-03-01T19:07:49.012780Z",
          "shell.execute_reply": "2021-03-01T19:07:49.013206Z"
        },
        "papermill": {
          "duration": 0.029767,
          "end_time": "2021-03-01T19:07:49.013347",
          "exception": false,
          "start_time": "2021-03-01T19:07:48.983580",
          "status": "completed"
        },
        "tags": [],
        "id": "3DNmku4kVFci"
      },
      "source": [
        "\n",
        "def clean_text(text):\n",
        "    text = text.lower()\n",
        "    text = re.sub(r\"what's\", \"what is \", text)\n",
        "    text = re.sub(r\"\\'s\", \" \", text)\n",
        "    text = re.sub(r\"\\'ve\", \" have \", text)\n",
        "    text = re.sub(r\"can't\", \"can not \", text)\n",
        "    text = re.sub(r\"n't\", \" not \", text)\n",
        "    text = re.sub(r\"i'm\", \"i am \", text)\n",
        "    text = re.sub(r\"\\'re\", \" are \", text)\n",
        "    text = re.sub(r\"\\'d\", \" would \", text)\n",
        "    text = re.sub(r\"\\'ll\", \" will \", text)\n",
        "    text = re.sub(r\"\\'scuse\", \" excuse \", text)\n",
        "    text = re.sub('\\W', ' ', text)\n",
        "    text = re.sub('\\s+', ' ', text)\n",
        "    text = text.strip(' ')\n",
        "    return text"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-03-01T19:07:49.059595Z",
          "iopub.status.busy": "2021-03-01T19:07:49.059039Z",
          "iopub.status.idle": "2021-03-01T19:08:01.968809Z",
          "shell.execute_reply": "2021-03-01T19:08:01.969576Z"
        },
        "papermill": {
          "duration": 12.936953,
          "end_time": "2021-03-01T19:08:01.969842",
          "exception": false,
          "start_time": "2021-03-01T19:07:49.032889",
          "status": "completed"
        },
        "tags": [],
        "id": "pef1imOUVFci"
      },
      "source": [
        "df['comment_text'] = df['comment_text'].map(lambda com : clean_text(com))"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-03-01T19:08:02.053174Z",
          "iopub.status.busy": "2021-03-01T19:08:02.052238Z",
          "iopub.status.idle": "2021-03-01T19:08:02.055780Z",
          "shell.execute_reply": "2021-03-01T19:08:02.053905Z"
        },
        "papermill": {
          "duration": 0.040371,
          "end_time": "2021-03-01T19:08:02.055939",
          "exception": false,
          "start_time": "2021-03-01T19:08:02.015568",
          "status": "completed"
        },
        "tags": [],
        "id": "0N004Va4VFcj",
        "outputId": "abf4cc8e-068e-4a90-c5d9-a42f2d2736cf"
      },
      "source": [
        "df['comment_text'][0]"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "'explanation why the edits made under my username hardcore metallica fan were reverted they were not vandalisms just closure on some gas after i voted at new york dolls fac and please do not remove the template from the talk page since i am retired now 89 205 38 27'"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 14
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-03-01T19:08:02.134947Z",
          "iopub.status.busy": "2021-03-01T19:08:02.134189Z",
          "iopub.status.idle": "2021-03-01T19:08:02.175968Z",
          "shell.execute_reply": "2021-03-01T19:08:02.177047Z"
        },
        "papermill": {
          "duration": 0.089236,
          "end_time": "2021-03-01T19:08:02.177252",
          "exception": false,
          "start_time": "2021-03-01T19:08:02.088016",
          "status": "completed"
        },
        "tags": [],
        "id": "7yli3roEVFcj"
      },
      "source": [
        "train, test = train_test_split(df, random_state=42, test_size=0.20, shuffle=True)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-03-01T19:08:02.259079Z",
          "iopub.status.busy": "2021-03-01T19:08:02.258181Z",
          "iopub.status.idle": "2021-03-01T19:08:02.262593Z",
          "shell.execute_reply": "2021-03-01T19:08:02.259804Z"
        },
        "papermill": {
          "duration": 0.049175,
          "end_time": "2021-03-01T19:08:02.262849",
          "exception": false,
          "start_time": "2021-03-01T19:08:02.213674",
          "status": "completed"
        },
        "tags": [],
        "id": "R0b6RrdyVFcj",
        "outputId": "fa30b5bb-ce13-4eff-8237-f07be86dbdb0"
      },
      "source": [
        "X_train = train.comment_text\n",
        "X_test = test.comment_text\n",
        "print(X_train.shape)\n",
        "print(X_test.shape)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "(127656,)\n",
            "(31915,)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-03-01T19:08:02.476729Z",
          "iopub.status.busy": "2021-03-01T19:08:02.475816Z",
          "iopub.status.idle": "2021-03-01T19:08:02.480390Z",
          "shell.execute_reply": "2021-03-01T19:08:02.483348Z"
        },
        "papermill": {
          "duration": 0.185909,
          "end_time": "2021-03-01T19:08:02.483824",
          "exception": false,
          "start_time": "2021-03-01T19:08:02.297915",
          "status": "completed"
        },
        "tags": [],
        "id": "RmlU36WnVFck",
        "outputId": "ec215fb3-4ff5-4e73-b200-033638de44f3"
      },
      "source": [
        "print(X_train)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "140030    grandma terri should burn in trash grandma ter...\n",
            "159124    9 may 2009 utc it would be easiest if you were...\n",
            "60006     the objectivity of this discussion is doubtful...\n",
            "65432                          shelly shock shelly shock is\n",
            "154979    i do not care refer to ong teng cheong talk pa...\n",
            "                                ...                        \n",
            "119879    redirect talk john loveday experimental physicist\n",
            "103694     back it up post the line here with the reference\n",
            "131932    i wo not stop that sometimes germanic equals g...\n",
            "146867    british bands i think you have mistaken scotti...\n",
            "121958    you are wrong justin thompson is mentioned in ...\n",
            "Name: comment_text, Length: 127656, dtype: object\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-03-01T19:08:02.560385Z",
          "iopub.status.busy": "2021-03-01T19:08:02.559702Z",
          "iopub.status.idle": "2021-03-01T19:08:02.572259Z",
          "shell.execute_reply": "2021-03-01T19:08:02.573187Z"
        },
        "papermill": {
          "duration": 0.055407,
          "end_time": "2021-03-01T19:08:02.573383",
          "exception": false,
          "start_time": "2021-03-01T19:08:02.517976",
          "status": "completed"
        },
        "tags": [],
        "id": "4ojR9kbHVFck",
        "outputId": "62f00948-f4d7-482d-db07-5a375a4ee6e3"
      },
      "source": [
        "y_train = train.drop(['id', 'comment_text'], axis=1)\n",
        "#y_train.columns = [''] * len(y_train.columns)\n",
        "y_test = test.drop(['id', 'comment_text'], axis=1)\n",
        "print(y_train)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "        toxic  severe_toxic  obscene  threat  insult  identity_hate  none\n",
            "140030      1             0        0       0       0              0     0\n",
            "159124      0             0        0       0       0              0     1\n",
            "60006       0             0        0       0       0              0     1\n",
            "65432       0             0        0       0       0              0     1\n",
            "154979      0             0        0       0       0              0     1\n",
            "...       ...           ...      ...     ...     ...            ...   ...\n",
            "119879      0             0        0       0       0              0     1\n",
            "103694      0             0        0       0       0              0     1\n",
            "131932      1             0        0       0       0              0     0\n",
            "146867      0             0        0       0       0              0     1\n",
            "121958      0             0        0       0       0              0     1\n",
            "\n",
            "[127656 rows x 7 columns]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-03-01T19:08:02.647547Z",
          "iopub.status.busy": "2021-03-01T19:08:02.646768Z",
          "iopub.status.idle": "2021-03-01T19:08:02.650406Z",
          "shell.execute_reply": "2021-03-01T19:08:02.650982Z"
        },
        "papermill": {
          "duration": 0.044949,
          "end_time": "2021-03-01T19:08:02.651179",
          "exception": false,
          "start_time": "2021-03-01T19:08:02.606230",
          "status": "completed"
        },
        "tags": [],
        "id": "qSzdRPl4VFck"
      },
      "source": [
        "NB_pipeline = Pipeline([\n",
        "                ('tfidf', TfidfVectorizer(stop_words=stop_words)),\n",
        "                ('clf', OneVsRestClassifier(MultinomialNB(\n",
        "                    fit_prior=True, class_prior=None, alpha=1e-3))),\n",
        "            ])"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-03-01T19:08:02.749815Z",
          "iopub.status.busy": "2021-03-01T19:08:02.724754Z",
          "iopub.status.idle": "2021-03-01T19:09:18.959916Z",
          "shell.execute_reply": "2021-03-01T19:09:18.960962Z"
        },
        "papermill": {
          "duration": 76.278957,
          "end_time": "2021-03-01T19:09:18.961214",
          "exception": false,
          "start_time": "2021-03-01T19:08:02.682257",
          "status": "completed"
        },
        "tags": [],
        "id": "CsH-03UmVFcl",
        "outputId": "0b53bb29-9596-45ca-fb80-608b8c4bc428"
      },
      "source": [
        "NB_pipeline = Pipeline([\n",
        "                ('tfidf', TfidfVectorizer(stop_words=stop_words)),\n",
        "                ('clf', OneVsRestClassifier(MultinomialNB(\n",
        "                    fit_prior=True, class_prior=None, alpha=1e-3))),\n",
        "            ])\n",
        "for category in categories:\n",
        "    print('... Processing {}'.format(category))\n",
        "    # train the model using X_dtm & y\n",
        "    NB_pipeline.fit(X_train, train[category])\n",
        "    # compute the testing accuracy\n",
        "    prediction = NB_pipeline.predict(X_test)\n",
        "    \n",
        "    \n",
        "    print('Test accuracy is {}'.format(accuracy_score(test[category], prediction)))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "... Processing toxic\n",
            "Test accuracy is 0.9434435218549272\n",
            "... Processing severe_toxic\n",
            "Test accuracy is 0.9893153689487701\n",
            "... Processing obscene\n",
            "Test accuracy is 0.9671001096663011\n",
            "... Processing threat\n",
            "Test accuracy is 0.9975246749177503\n",
            "... Processing insult\n",
            "Test accuracy is 0.9639041203195989\n",
            "... Processing identity_hate\n",
            "Test accuracy is 0.9898167006109979\n",
            "... Processing none\n",
            "Test accuracy is 0.9417515274949083\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-03-01T19:09:19.055144Z",
          "iopub.status.busy": "2021-03-01T19:09:19.052881Z",
          "iopub.status.idle": "2021-03-01T19:09:19.056255Z",
          "shell.execute_reply": "2021-03-01T19:09:19.053994Z"
        },
        "papermill": {
          "duration": 0.05161,
          "end_time": "2021-03-01T19:09:19.056509",
          "exception": false,
          "start_time": "2021-03-01T19:09:19.004899",
          "status": "completed"
        },
        "tags": [],
        "id": "ZVxYWqLVVFcl"
      },
      "source": [
        "\n",
        "SVC_pipeline = Pipeline([\n",
        "                ('tfidf', TfidfVectorizer(stop_words=stop_words)),\n",
        "                ('clf', OneVsRestClassifier(LinearSVC(), n_jobs=1)),\n",
        "            ])"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-03-01T19:09:19.160502Z",
          "iopub.status.busy": "2021-03-01T19:09:19.159726Z",
          "iopub.status.idle": "2021-03-01T19:10:41.199460Z",
          "shell.execute_reply": "2021-03-01T19:10:41.200064Z"
        },
        "papermill": {
          "duration": 82.092762,
          "end_time": "2021-03-01T19:10:41.200277",
          "exception": false,
          "start_time": "2021-03-01T19:09:19.107515",
          "status": "completed"
        },
        "tags": [],
        "id": "45UDfxrWVFcl",
        "outputId": "cd2b9be8-a19b-4c68-a306-4f6fba1b94db"
      },
      "source": [
        "SVC_pipeline = Pipeline([\n",
        "                ('tfidf', TfidfVectorizer(stop_words=stop_words)),\n",
        "                ('clf', OneVsRestClassifier(LinearSVC(), n_jobs=1)),\n",
        "            ])\n",
        "for category in categories:\n",
        "    print('... Processing {}'.format(category))\n",
        "    # train the model using X_dtm & y\n",
        "    SVC_pipeline.fit(X_train, train[category])\n",
        "    # compute the testing accuracy\n",
        "    prediction = SVC_pipeline.predict(X_test)\n",
        "    print('Test accuracy is {}'.format(accuracy_score(test[category], prediction)))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "... Processing toxic\n",
            "Test accuracy is 0.9602694657684474\n",
            "... Processing severe_toxic\n",
            "Test accuracy is 0.9906626977910074\n",
            "... Processing obscene\n",
            "Test accuracy is 0.9794454018486605\n",
            "... Processing threat\n",
            "Test accuracy is 0.9976813410621964\n",
            "... Processing insult\n",
            "Test accuracy is 0.9711107629641235\n",
            "... Processing identity_hate\n",
            "Test accuracy is 0.9920726930910231\n",
            "... Processing none\n",
            "Test accuracy is 0.9606767977440075\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-03-01T19:10:41.310026Z",
          "iopub.status.busy": "2021-03-01T19:10:41.294585Z",
          "iopub.status.idle": "2021-03-01T19:12:19.634812Z",
          "shell.execute_reply": "2021-03-01T19:12:19.635348Z"
        },
        "papermill": {
          "duration": 98.40842,
          "end_time": "2021-03-01T19:12:19.635527",
          "exception": false,
          "start_time": "2021-03-01T19:10:41.227107",
          "status": "completed"
        },
        "tags": [],
        "id": "sQT9b0EEVFcm",
        "outputId": "61fc4c15-854e-4bf1-fcb2-c5154e3fecd1"
      },
      "source": [
        "LogReg_pipeline = Pipeline([\n",
        "                ('tfidf', TfidfVectorizer(stop_words=stop_words)),\n",
        "                ('clf', OneVsRestClassifier(LogisticRegression(solver='sag', C=7), n_jobs=1)),\n",
        "            ])\n",
        "for category in categories:\n",
        "    print('... Processing {}'.format(category))\n",
        "    # train the model using X_dtm & y\n",
        "    LogReg_pipeline.fit(X_train, train[category])\n",
        "    # compute the testing accuracy\n",
        "    prediction = LogReg_pipeline.predict(X_test)\n",
        "    print('Test accuracy is {}'.format(accuracy_score(test[category], prediction)))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "... Processing toxic\n",
            "Test accuracy is 0.9601127996240012\n",
            "... Processing severe_toxic\n",
            "Test accuracy is 0.9907253642487859\n",
            "... Processing obscene\n",
            "Test accuracy is 0.9791947360175466\n",
            "... Processing threat\n",
            "Test accuracy is 0.9976500078333073\n",
            "... Processing insult\n",
            "Test accuracy is 0.9708600971330096\n",
            "... Processing identity_hate\n",
            "Test accuracy is 0.9918846937176876\n",
            "... Processing none\n",
            "Test accuracy is 0.9603007989973367\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-03-01T19:12:19.697027Z",
          "iopub.status.busy": "2021-03-01T19:12:19.696329Z",
          "iopub.status.idle": "2021-03-01T19:12:19.811553Z",
          "shell.execute_reply": "2021-03-01T19:12:19.811023Z"
        },
        "papermill": {
          "duration": 0.147615,
          "end_time": "2021-03-01T19:12:19.811690",
          "exception": false,
          "start_time": "2021-03-01T19:12:19.664075",
          "status": "completed"
        },
        "tags": [],
        "id": "ZpRc18fAVFcm"
      },
      "source": [
        "from sklearn.tree import DecisionTreeClassifier\n",
        "from sklearn.ensemble import RandomForestClassifier"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-03-01T19:12:19.921049Z",
          "iopub.status.busy": "2021-03-01T19:12:19.910827Z",
          "iopub.status.idle": "2021-03-01T19:13:52.398707Z",
          "shell.execute_reply": "2021-03-01T19:13:52.399239Z"
        },
        "papermill": {
          "duration": 92.559164,
          "end_time": "2021-03-01T19:13:52.399420",
          "exception": false,
          "start_time": "2021-03-01T19:12:19.840256",
          "status": "completed"
        },
        "tags": [],
        "id": "AQFpiMriVFcn",
        "outputId": "67ab9e3d-74db-492c-9c11-21dc80917865"
      },
      "source": [
        "RanFor_pipeline = Pipeline([\n",
        "                ('tfidf', TfidfVectorizer(stop_words=stop_words)),\n",
        "                ('clf', OneVsRestClassifier(RandomForestClassifier(max_depth= 15, n_estimators=40), n_jobs=1)),\n",
        "            ])\n",
        "for category in categories:\n",
        "    print('... Processing {}'.format(category))\n",
        "    # train the model using X_dtm & y\n",
        "    RanFor_pipeline.fit(X_train, train[category])\n",
        "    # compute the testing accuracy\n",
        "    prediction = RanFor_pipeline.predict(X_test)\n",
        "    print('Test accuracy is {}'.format(accuracy_score(test[category], prediction)))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "... Processing toxic\n",
            "Test accuracy is 0.9042456525144916\n",
            "... Processing severe_toxic\n",
            "Test accuracy is 0.989942033526555\n",
            "... Processing obscene\n",
            "Test accuracy is 0.9462635124549584\n",
            "... Processing threat\n",
            "Test accuracy is 0.9976813410621964\n",
            "... Processing insult\n",
            "Test accuracy is 0.9494281685727715\n",
            "... Processing identity_hate\n",
            "Test accuracy is 0.9907880307065643\n",
            "... Processing none\n",
            "Test accuracy is 0.8983863387122043\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-03-01T19:13:52.507411Z",
          "iopub.status.busy": "2021-03-01T19:13:52.502495Z",
          "iopub.status.idle": "2021-03-01T19:16:52.819481Z",
          "shell.execute_reply": "2021-03-01T19:16:52.820078Z"
        },
        "papermill": {
          "duration": 180.390046,
          "end_time": "2021-03-01T19:16:52.820283",
          "exception": false,
          "start_time": "2021-03-01T19:13:52.430237",
          "status": "completed"
        },
        "tags": [],
        "id": "xaRQUkH6VFcn",
        "outputId": "b9cd9af0-ff43-4175-f6ac-b88649753b70"
      },
      "source": [
        "DecTree_pipeline = Pipeline([\n",
        "                ('tfidf', TfidfVectorizer(stop_words=stop_words)),\n",
        "                ('clf', OneVsRestClassifier(DecisionTreeClassifier(max_depth = 25), n_jobs=1)),\n",
        "            ])\n",
        "for category in categories:\n",
        "    print('... Processing {}'.format(category))\n",
        "    # train the model using X_dtm & y\n",
        "    DecTree_pipeline.fit(X_train, train[category])\n",
        "    # compute the testing accuracy\n",
        "    prediction = DecTree_pipeline.predict(X_test)\n",
        "    print('Test accuracy is {}'.format(accuracy_score(test[category], prediction)))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "... Processing toxic\n",
            "Test accuracy is 0.9460128466238445\n",
            "... Processing severe_toxic\n",
            "Test accuracy is 0.9884380385398716\n",
            "... Processing obscene\n",
            "Test accuracy is 0.9772834090553032\n",
            "... Processing threat\n",
            "Test accuracy is 0.9973053423155256\n",
            "... Processing insult\n",
            "Test accuracy is 0.9652827823907254\n",
            "... Processing identity_hate\n",
            "Test accuracy is 0.991665361115463\n",
            "... Processing none\n",
            "Test accuracy is 0.944477518408272\n"
          ],
          "name": "stdout"
        }
      ]
    }
  ]
}