Documentation.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "65b8ad8c",
   "metadata": {},
   "source": [
    "# How to Run Interface"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "16955d3a",
   "metadata": {},
   "source": [
    "1. first run this on terminal/bash on your computer using\n",
    "    ``pip install -r requirements.txt`` script\n",
    "2. Then run the model using script\n",
    "    ``python3 model.py``"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "70efb418",
   "metadata": {},
   "source": [
    "# Documentation of Modelling process"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 216,
   "id": "865312d1",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-11-25T08:01:49.988090Z",
     "start_time": "2021-11-25T08:01:49.980387Z"
    }
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import re\n",
    "import matplotlib.pyplot as plt\n",
    "import pickle\n",
    "import seaborn as sns\n",
    "import numpy as np\n",
    "np.random.seed(1718)\n",
    "import warnings\n",
    "warnings.filterwarnings(\"ignore\")\n",
    "\n",
    "np.random.seed(0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 217,
   "id": "64c56bd0",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-11-25T08:01:50.524694Z",
     "start_time": "2021-11-25T08:01:49.991342Z"
    }
   },
   "outputs": [],
   "source": [
    "df = pd.read_csv(\"data.csv\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 218,
   "id": "e00d71ef",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-11-25T08:01:50.550919Z",
     "start_time": "2021-11-25T08:01:50.527941Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 10000 entries, 0 to 9999\n",
      "Data columns (total 3 columns):\n",
      " #   Column           Non-Null Count  Dtype \n",
      "---  ------           --------------  ----- \n",
      " 0   article_id       10000 non-null  int64 \n",
      " 1   article_topic    10000 non-null  object\n",
      " 2   article_content  9964 non-null   object\n",
      "dtypes: int64(1), object(2)\n",
      "memory usage: 234.5+ KB\n"
     ]
    }
   ],
   "source": [
    "df.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ca341693",
   "metadata": {},
   "source": [
    "# Data Exploration"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 219,
   "id": "dbc27db3",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-11-25T08:01:50.564399Z",
     "start_time": "2021-11-25T08:01:50.554183Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "article_id          0\n",
       "article_topic       0\n",
       "article_content    36\n",
       "dtype: int64"
      ]
     },
     "execution_count": 219,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.isna().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 220,
   "id": "22f8a158",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-11-25T08:01:50.584869Z",
     "start_time": "2021-11-25T08:01:50.566875Z"
    }
   },
   "outputs": [
    {
     "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_id</th>\n",
       "      <th>article_topic</th>\n",
       "      <th>article_content</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>197</th>\n",
       "      <td>93210288</td>\n",
       "      <td>Teknologi</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>674</th>\n",
       "      <td>93185319</td>\n",
       "      <td>Hiburan</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>817</th>\n",
       "      <td>93189481</td>\n",
       "      <td>Hiburan</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>972</th>\n",
       "      <td>93184085</td>\n",
       "      <td>Otomotif</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015</th>\n",
       "      <td>93195291</td>\n",
       "      <td>Hiburan</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2250</th>\n",
       "      <td>93201544</td>\n",
       "      <td>Hiburan</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3276</th>\n",
       "      <td>93213891</td>\n",
       "      <td>Teknologi</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4150</th>\n",
       "      <td>93197166</td>\n",
       "      <td>Hiburan</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4338</th>\n",
       "      <td>93186717</td>\n",
       "      <td>Sepak Bola</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4750</th>\n",
       "      <td>93224988</td>\n",
       "      <td>Sepak Bola</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4838</th>\n",
       "      <td>93183169</td>\n",
       "      <td>Politik</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4917</th>\n",
       "      <td>93194456</td>\n",
       "      <td>Hiburan</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5126</th>\n",
       "      <td>93184755</td>\n",
       "      <td>Hiburan</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5306</th>\n",
       "      <td>93195757</td>\n",
       "      <td>Lifestyle</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5859</th>\n",
       "      <td>93208751</td>\n",
       "      <td>Hiburan</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5965</th>\n",
       "      <td>93189109</td>\n",
       "      <td>Lifestyle</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6032</th>\n",
       "      <td>93201441</td>\n",
       "      <td>Internasional</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6094</th>\n",
       "      <td>93187589</td>\n",
       "      <td>Hiburan</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6442</th>\n",
       "      <td>947838</td>\n",
       "      <td>Ekonomi</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6539</th>\n",
       "      <td>93181264</td>\n",
       "      <td>Lifestyle</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6689</th>\n",
       "      <td>845044</td>\n",
       "      <td>Hiburan</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6760</th>\n",
       "      <td>93188937</td>\n",
       "      <td>Hiburan</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6816</th>\n",
       "      <td>93195181</td>\n",
       "      <td>Internasional</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6886</th>\n",
       "      <td>93201878</td>\n",
       "      <td>Sepak Bola</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7239</th>\n",
       "      <td>93187815</td>\n",
       "      <td>Hiburan</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7307</th>\n",
       "      <td>93182251</td>\n",
       "      <td>Hiburan</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7403</th>\n",
       "      <td>93186538</td>\n",
       "      <td>Hiburan</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7425</th>\n",
       "      <td>93186516</td>\n",
       "      <td>Sepak Bola</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7490</th>\n",
       "      <td>93192172</td>\n",
       "      <td>Teknologi</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7526</th>\n",
       "      <td>93203404</td>\n",
       "      <td>Hiburan</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7920</th>\n",
       "      <td>93208905</td>\n",
       "      <td>Hiburan</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8457</th>\n",
       "      <td>93201720</td>\n",
       "      <td>Teknologi</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8833</th>\n",
       "      <td>93189224</td>\n",
       "      <td>Ekonomi</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8851</th>\n",
       "      <td>93191464</td>\n",
       "      <td>Lifestyle</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9552</th>\n",
       "      <td>93183963</td>\n",
       "      <td>Hiburan</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9696</th>\n",
       "      <td>93201912</td>\n",
       "      <td>Hiburan</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      article_id  article_topic article_content\n",
       "197     93210288      Teknologi             NaN\n",
       "674     93185319        Hiburan             NaN\n",
       "817     93189481        Hiburan             NaN\n",
       "972     93184085       Otomotif             NaN\n",
       "2015    93195291        Hiburan             NaN\n",
       "2250    93201544        Hiburan             NaN\n",
       "3276    93213891      Teknologi             NaN\n",
       "4150    93197166        Hiburan             NaN\n",
       "4338    93186717     Sepak Bola             NaN\n",
       "4750    93224988     Sepak Bola             NaN\n",
       "4838    93183169        Politik             NaN\n",
       "4917    93194456        Hiburan             NaN\n",
       "5126    93184755        Hiburan             NaN\n",
       "5306    93195757      Lifestyle             NaN\n",
       "5859    93208751        Hiburan             NaN\n",
       "5965    93189109      Lifestyle             NaN\n",
       "6032    93201441  Internasional             NaN\n",
       "6094    93187589        Hiburan             NaN\n",
       "6442      947838        Ekonomi             NaN\n",
       "6539    93181264      Lifestyle             NaN\n",
       "6689      845044        Hiburan             NaN\n",
       "6760    93188937        Hiburan             NaN\n",
       "6816    93195181  Internasional             NaN\n",
       "6886    93201878     Sepak Bola             NaN\n",
       "7239    93187815        Hiburan             NaN\n",
       "7307    93182251        Hiburan             NaN\n",
       "7403    93186538        Hiburan             NaN\n",
       "7425    93186516     Sepak Bola             NaN\n",
       "7490    93192172      Teknologi             NaN\n",
       "7526    93203404        Hiburan             NaN\n",
       "7920    93208905        Hiburan             NaN\n",
       "8457    93201720      Teknologi             NaN\n",
       "8833    93189224        Ekonomi             NaN\n",
       "8851    93191464      Lifestyle             NaN\n",
       "9552    93183963        Hiburan             NaN\n",
       "9696    93201912        Hiburan             NaN"
      ]
     },
     "execution_count": 220,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[df.isnull().any(axis=1)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 221,
   "id": "e0369388",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-11-25T08:01:50.605767Z",
     "start_time": "2021-11-25T08:01:50.589695Z"
    }
   },
   "outputs": [],
   "source": [
    "df = df.dropna()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 222,
   "id": "91530c12",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-11-25T08:01:51.230903Z",
     "start_time": "2021-11-25T08:01:50.608903Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 792x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "df['content_length'] = [len(x.split()) for x in df['article_content'].tolist()]\n",
    "plt.figure(figsize=(11,6))\n",
    "sns.distplot(df['content_length']).set_title('article_content distribution based on content lenght');\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 223,
   "id": "fc36d1fe",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-11-25T08:01:51.246007Z",
     "start_time": "2021-11-25T08:01:51.233849Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count    9964.000000\n",
       "mean      353.124548\n",
       "std       297.024979\n",
       "min         0.000000\n",
       "25%       233.000000\n",
       "50%       303.000000\n",
       "75%       400.000000\n",
       "max      9921.000000\n",
       "Name: content_length, dtype: float64"
      ]
     },
     "execution_count": 223,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['content_length'].describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 224,
   "id": "3a4a8d44",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-11-25T08:01:51.269300Z",
     "start_time": "2021-11-25T08:01:51.249236Z"
    }
   },
   "outputs": [
    {
     "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_id</th>\n",
       "      <th>article_topic</th>\n",
       "      <th>article_content</th>\n",
       "      <th>content_length</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>418</th>\n",
       "      <td>93185667</td>\n",
       "      <td>Haji</td>\n",
       "      <td>tes</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>555</th>\n",
       "      <td>93181418</td>\n",
       "      <td>Hiburan</td>\n",
       "      <td></td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>971</th>\n",
       "      <td>93204251</td>\n",
       "      <td>Teknologi</td>\n",
       "      <td>bisa dilihat disini, hak cipta bukan milik saya</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1331</th>\n",
       "      <td>93191657</td>\n",
       "      <td>Hiburan</td>\n",
       "      <td>,</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1432</th>\n",
       "      <td>93212920</td>\n",
       "      <td>Kesehatan</td>\n",
       "      <td></td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1568</th>\n",
       "      <td>93198989</td>\n",
       "      <td>Sepak Bola</td>\n",
       "      <td>test</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1582</th>\n",
       "      <td>93182448</td>\n",
       "      <td>Hukum</td>\n",
       "      <td>Panjang bener!</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1601</th>\n",
       "      <td>93181411</td>\n",
       "      <td>Hiburan</td>\n",
       "      <td></td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1927</th>\n",
       "      <td>93185665</td>\n",
       "      <td>Haji</td>\n",
       "      <td>tes</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2031</th>\n",
       "      <td>93210296</td>\n",
       "      <td>Health</td>\n",
       "      <td>.,</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2342</th>\n",
       "      <td>93190978</td>\n",
       "      <td>Hiburan</td>\n",
       "      <td></td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2937</th>\n",
       "      <td>93190761</td>\n",
       "      <td>Hiburan</td>\n",
       "      <td>.</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3186</th>\n",
       "      <td>93184385</td>\n",
       "      <td>Sepak Bola</td>\n",
       "      <td>Selamat datang di SportToday, pusatnya informa...</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3600</th>\n",
       "      <td>1485824</td>\n",
       "      <td>Hiburan</td>\n",
       "      <td></td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3730</th>\n",
       "      <td>93191661</td>\n",
       "      <td>Hiburan</td>\n",
       "      <td>,</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4085</th>\n",
       "      <td>1586004</td>\n",
       "      <td>Politik</td>\n",
       "      <td></td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4087</th>\n",
       "      <td>4052810</td>\n",
       "      <td>Politik</td>\n",
       "      <td></td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4173</th>\n",
       "      <td>93191699</td>\n",
       "      <td>Hiburan</td>\n",
       "      <td>.</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5297</th>\n",
       "      <td>93216759</td>\n",
       "      <td>Kesehatan</td>\n",
       "      <td></td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5460</th>\n",
       "      <td>93189565</td>\n",
       "      <td>Hiburan</td>\n",
       "      <td>....No comment....</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6352</th>\n",
       "      <td>1484488</td>\n",
       "      <td>Sepak Bola</td>\n",
       "      <td></td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6887</th>\n",
       "      <td>93181415</td>\n",
       "      <td>Hiburan</td>\n",
       "      <td></td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6942</th>\n",
       "      <td>93188749</td>\n",
       "      <td>Politik</td>\n",
       "      <td></td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7180</th>\n",
       "      <td>93200472</td>\n",
       "      <td>Politik</td>\n",
       "      <td>Inspiring quotes</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7305</th>\n",
       "      <td>93182451</td>\n",
       "      <td>Otomotif</td>\n",
       "      <td>Ini gaya gue!</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8074</th>\n",
       "      <td>93190979</td>\n",
       "      <td>Hiburan</td>\n",
       "      <td></td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8351</th>\n",
       "      <td>93234476</td>\n",
       "      <td>Travel</td>\n",
       "      <td>Selong Belanak Lombok Indonesia</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8697</th>\n",
       "      <td>93195225</td>\n",
       "      <td>Kesehatan</td>\n",
       "      <td></td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9120</th>\n",
       "      <td>93189058</td>\n",
       "      <td>Hiburan</td>\n",
       "      <td>Suaranya Imut-imut Lucu.</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9261</th>\n",
       "      <td>93191674</td>\n",
       "      <td>Hiburan</td>\n",
       "      <td>.</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9375</th>\n",
       "      <td>93228174</td>\n",
       "      <td>Personal</td>\n",
       "      <td>HellojikaAhshhshdhdjjdshhdjdjjdjdjjd</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9899</th>\n",
       "      <td>93228169</td>\n",
       "      <td>Personal</td>\n",
       "      <td>Hello</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      article_id article_topic  \\\n",
       "418     93185667          Haji   \n",
       "555     93181418       Hiburan   \n",
       "971     93204251     Teknologi   \n",
       "1331    93191657       Hiburan   \n",
       "1432    93212920     Kesehatan   \n",
       "1568    93198989    Sepak Bola   \n",
       "1582    93182448         Hukum   \n",
       "1601    93181411       Hiburan   \n",
       "1927    93185665          Haji   \n",
       "2031    93210296        Health   \n",
       "2342    93190978       Hiburan   \n",
       "2937    93190761       Hiburan   \n",
       "3186    93184385    Sepak Bola   \n",
       "3600     1485824       Hiburan   \n",
       "3730    93191661       Hiburan   \n",
       "4085     1586004       Politik   \n",
       "4087     4052810       Politik   \n",
       "4173    93191699       Hiburan   \n",
       "5297    93216759     Kesehatan   \n",
       "5460    93189565       Hiburan   \n",
       "6352     1484488    Sepak Bola   \n",
       "6887    93181415       Hiburan   \n",
       "6942    93188749       Politik   \n",
       "7180    93200472       Politik   \n",
       "7305    93182451      Otomotif   \n",
       "8074    93190979       Hiburan   \n",
       "8351    93234476        Travel   \n",
       "8697    93195225     Kesehatan   \n",
       "9120    93189058       Hiburan   \n",
       "9261    93191674       Hiburan   \n",
       "9375    93228174      Personal   \n",
       "9899    93228169      Personal   \n",
       "\n",
       "                                        article_content  content_length  \n",
       "418                                                tes                1  \n",
       "555                                                                   0  \n",
       "971    bisa dilihat disini, hak cipta bukan milik saya                8  \n",
       "1331                                                  ,               1  \n",
       "1432                                                                  0  \n",
       "1568                                               test               1  \n",
       "1582                                    Panjang bener!                2  \n",
       "1601                                                                  0  \n",
       "1927                                                tes               1  \n",
       "2031                                               .,                 1  \n",
       "2342                                                                  0  \n",
       "2937                                                  .               1  \n",
       "3186  Selamat datang di SportToday, pusatnya informa...               9  \n",
       "3600                                                                  0  \n",
       "3730                                                  ,               1  \n",
       "4085                                                                  0  \n",
       "4087                                                                  0  \n",
       "4173                                                  .               1  \n",
       "5297                                                                  0  \n",
       "5460                                 ....No comment....               2  \n",
       "6352                                                                  0  \n",
       "6887                                                                  0  \n",
       "6942                                                                  0  \n",
       "7180                                 Inspiring quotes                 2  \n",
       "7305                                     Ini gaya gue!                3  \n",
       "8074                                                                  0  \n",
       "8351                   Selong Belanak Lombok Indonesia                4  \n",
       "8697                                                                  0  \n",
       "9120                          Suaranya Imut-imut Lucu.                3  \n",
       "9261                                                  .               1  \n",
       "9375              HellojikaAhshhshdhdjjdshhdjdjjdjdjjd                1  \n",
       "9899                                             Hello                1  "
      ]
     },
     "execution_count": 224,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[df['content_length'] < 10]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 225,
   "id": "56e9ed86",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-11-25T08:01:51.278237Z",
     "start_time": "2021-11-25T08:01:51.271556Z"
    }
   },
   "outputs": [],
   "source": [
    "df.drop(df[df['content_length'] < 5].index, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 226,
   "id": "2eb3fd1f",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-11-25T08:01:51.301060Z",
     "start_time": "2021-11-25T08:01:51.281654Z"
    }
   },
   "outputs": [],
   "source": [
    "df = df.reset_index(drop=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 227,
   "id": "c9fc4f2e",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-11-25T08:01:51.636202Z",
     "start_time": "2021-11-25T08:01:51.304533Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "df['article_topic'].value_counts().plot(figsize=(10, 5), kind='bar', title='Content Topic Distribution')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 228,
   "id": "bf672129",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-11-25T08:01:51.647910Z",
     "start_time": "2021-11-25T08:01:51.639121Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count    9934.000000\n",
       "mean      354.188242\n",
       "std       296.840804\n",
       "min         8.000000\n",
       "25%       234.000000\n",
       "50%       303.000000\n",
       "75%       400.000000\n",
       "max      9921.000000\n",
       "Name: content_length, dtype: float64"
      ]
     },
     "execution_count": 228,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['content_length'].describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 229,
   "id": "2e40e16f",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-11-25T08:01:51.658289Z",
     "start_time": "2021-11-25T08:01:51.650038Z"
    }
   },
   "outputs": [
    {
     "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_id</th>\n",
       "      <th>article_topic</th>\n",
       "      <th>article_content</th>\n",
       "      <th>content_length</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>93205794</td>\n",
       "      <td>Internasional</td>\n",
       "      <td>Kepolisian Inggris tengah memburu pelaku yang...</td>\n",
       "      <td>164</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>93186698</td>\n",
       "      <td>Ekonomi</td>\n",
       "      <td>Seluruh layanan transaksi di jalan tol akan m...</td>\n",
       "      <td>373</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>93191463</td>\n",
       "      <td>Teknologi</td>\n",
       "      <td>\\nHari ini, Rabu (23/8), ternyata menjadi har...</td>\n",
       "      <td>165</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>93219292</td>\n",
       "      <td>Ekonomi</td>\n",
       "      <td>Saat ini Indonesia hanya memiliki cadangan ba...</td>\n",
       "      <td>352</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>343106</td>\n",
       "      <td>Hiburan</td>\n",
       "      <td>Hari ini, Selasa (1/8), pedangdut Ridho Rhoma...</td>\n",
       "      <td>281</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   article_id  article_topic  \\\n",
       "0    93205794  Internasional   \n",
       "1    93186698        Ekonomi   \n",
       "2    93191463      Teknologi   \n",
       "3    93219292        Ekonomi   \n",
       "4      343106        Hiburan   \n",
       "\n",
       "                                     article_content  content_length  \n",
       "0   Kepolisian Inggris tengah memburu pelaku yang...             164  \n",
       "1   Seluruh layanan transaksi di jalan tol akan m...             373  \n",
       "2   \\nHari ini, Rabu (23/8), ternyata menjadi har...             165  \n",
       "3   Saat ini Indonesia hanya memiliki cadangan ba...             352  \n",
       "4   Hari ini, Selasa (1/8), pedangdut Ridho Rhoma...             281  "
      ]
     },
     "execution_count": 229,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9db886b0",
   "metadata": {},
   "source": [
    "Reports :\n",
    "- There are 36 null data (Done)\n",
    "- There is empty text and no content in it (Done)\n",
    "- There is very long content reaching 9000 words but the average content contains approximatly 350 words\n",
    "- There are 29 Class Targets\n",
    "- There is an imbalance in the data with data dominated by the topics of Economy, Hajj, Entertainment, Football and International\n",
    "- There are classes that look like they can be combined like 'health' and 'health' and 'k-pop' with 'entertainment but I'll leave the target class as it is"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c74b8a56",
   "metadata": {},
   "source": [
    "# Text Preprocessing and feature engineering for model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 230,
   "id": "5d7cd1cc",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-11-25T08:01:51.680168Z",
     "start_time": "2021-11-25T08:01:51.660680Z"
    }
   },
   "outputs": [],
   "source": [
    "df['labels'] = df['article_topic'].factorize()[0]\n",
    "from io import StringIO\n",
    "category_id_df = df[['article_topic', 'labels']].drop_duplicates().sort_values('labels')\n",
    "category_to_id = dict(category_id_df.values)\n",
    "id_to_category = dict(category_id_df[['labels', 'article_topic']].values)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 231,
   "id": "fcfeb27e",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-11-25T08:01:51.687040Z",
     "start_time": "2021-11-25T08:01:51.682995Z"
    }
   },
   "outputs": [],
   "source": [
    "def print_text(index):\n",
    "    example = df[df.index == index][['article_content', 'article_topic']].values[0]\n",
    "    if len(example) > 0:\n",
    "        print(example[0])\n",
    "        print('article_topic:', example[1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 232,
   "id": "cb5ae015",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-11-25T08:01:51.693133Z",
     "start_time": "2021-11-25T08:01:51.688773Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " \n",
      "Hari ini, Rabu (23/8), ternyata menjadi hari yang spesial bagi tagar atau yang biasa disebut juga hashtag. Kenapa demikian? 23 Agustus ternyata menandai tepat 10 tahun pasca diciptakannya tagar (#) di media sosial Twitter.\n",
      "\n",
      "Untuk merayakan hari ulang tahun ke-10 dari tagar, Twitter pun menghadirkan emoji khusus yang akan keluar ketika kita mengetikkan #Hashtag10 di platform media sosial tersebut.\n",
      " \n",
      "Twitter memang menjadi platform media sosial yang mempopulerkan tagar, di mana tagar ini dapat menggabungkan seluruh kicauan dalam suatu topik yang sama ketika pengguna mengklik tagar tersebut.\n",
      "\n",
      "Selain itu, Twitter juga merilis daftar 10 tagar yang paling populer yang digunakan oleh para penggunanya di Indonesia selama 10 tahun terakhir. Penasaran apa saja? Berikut daftarnya.\n",
      "\n",
      "1. #OwiButetBisaEmas2. #Salam2Jari3. #KamiTidakTakut4. #SaveKPK5. #Euro20166. #PersibDay7. #MUFC8. #MelawanAsap9. #DemiMetallica10. #JumatBerkah\n",
      " \n",
      "Dari deretan tagar terpopuler ini terlihat bagaimana animo masyarakat menyebar ke berbagai topik yang bermacam-macam. Ada tagar mengenai politik, olahraga, konser Metallica di Jakarta, hingga sebuah pernyataan seperti #KamiTidakTakut yang bergema pasca serangan teroris di Tanah Air.\n",
      "\n",
      "Jadi, mana tagar favoritmu?\n",
      " \n",
      "article_topic: Teknologi\n"
     ]
    }
   ],
   "source": [
    "print_text(2)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e827ab09",
   "metadata": {},
   "source": [
    "##### After looking through the contents of some content, I think, I don't need to clean up too complex text just work with symbols, numbers and capital letters"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 233,
   "id": "370e825b",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-11-25T08:01:51.699230Z",
     "start_time": "2021-11-25T08:01:51.695452Z"
    }
   },
   "outputs": [],
   "source": [
    "from nlp_id.stopword import StopWord \n",
    "from nlp_id.lemmatizer import Lemmatizer "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 286,
   "id": "dcd8b652",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-11-27T04:01:33.372537Z",
     "start_time": "2021-11-27T04:01:33.363996Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1168"
      ]
     },
     "execution_count": 286,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(stopword.get_stopword()) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 234,
   "id": "3a86d93c",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-11-25T08:01:52.824766Z",
     "start_time": "2021-11-25T08:01:51.701498Z"
    }
   },
   "outputs": [],
   "source": [
    "lemmatizer = Lemmatizer() \n",
    "stopword = StopWord()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 235,
   "id": "2dfc187f",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-11-25T08:02:37.749198Z",
     "start_time": "2021-11-25T08:01:52.827737Z"
    }
   },
   "outputs": [],
   "source": [
    "corpus = []\n",
    "for i in range(0, len(df)):\n",
    "    review = re.sub('[^a-zA-Z]', ' ', df['article_content'][i])\n",
    "    review = lemmatizer.lemmatize(review)\n",
    "    review = stopword.remove_stopword(review)\n",
    "    review = review.lower()\n",
    "    review = review.split()\n",
    "    review = ' '.join(review)\n",
    "    corpus.append(review)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 236,
   "id": "9ab0a257",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-11-25T08:02:37.758717Z",
     "start_time": "2021-11-25T08:02:37.751541Z"
    }
   },
   "outputs": [],
   "source": [
    "df['corpus'] = corpus"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 237,
   "id": "d955d2d5",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-11-25T08:02:37.775487Z",
     "start_time": "2021-11-25T08:02:37.761059Z"
    }
   },
   "outputs": [
    {
     "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_id</th>\n",
       "      <th>article_topic</th>\n",
       "      <th>article_content</th>\n",
       "      <th>content_length</th>\n",
       "      <th>labels</th>\n",
       "      <th>corpus</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>3493</th>\n",
       "      <td>93212446</td>\n",
       "      <td>Hiburan</td>\n",
       "      <td>Penyanyi muda Isyana Sarasvati secara resmi m...</td>\n",
       "      <td>391</td>\n",
       "      <td>3</td>\n",
       "      <td>nyanyi muda isyana sarasvati resmi luncur albu...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1785</th>\n",
       "      <td>93197552</td>\n",
       "      <td>Sepak Bola</td>\n",
       "      <td>Kembalinya Tim Nasional Indonesia senior ke m...</td>\n",
       "      <td>592</td>\n",
       "      <td>7</td>\n",
       "      <td>tim nasional indonesia senior medan laga kecew...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8633</th>\n",
       "      <td>93210133</td>\n",
       "      <td>Bojonegoro</td>\n",
       "      <td>Reporter: Parto SasmitoblokBojonegoro.com - S...</td>\n",
       "      <td>153</td>\n",
       "      <td>20</td>\n",
       "      <td>reporter parto sasmitoblokbojonegoro com senam...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8625</th>\n",
       "      <td>93226521</td>\n",
       "      <td>Ekonomi</td>\n",
       "      <td>Perusahaan BUMN konstruksi, PT Adhi Karya Tbk...</td>\n",
       "      <td>220</td>\n",
       "      <td>1</td>\n",
       "      <td>usaha bumn konstruksi pt adhi karya tbk inside...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4906</th>\n",
       "      <td>93185131</td>\n",
       "      <td>Ekonomi</td>\n",
       "      <td>Ekonom Universitas Indonesia (UI) Faisal Basr...</td>\n",
       "      <td>342</td>\n",
       "      <td>1</td>\n",
       "      <td>ekonom universitas indonesia ui faisal basri d...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      article_id article_topic  \\\n",
       "3493    93212446       Hiburan   \n",
       "1785    93197552    Sepak Bola   \n",
       "8633    93210133    Bojonegoro   \n",
       "8625    93226521       Ekonomi   \n",
       "4906    93185131       Ekonomi   \n",
       "\n",
       "                                        article_content  content_length  \\\n",
       "3493   Penyanyi muda Isyana Sarasvati secara resmi m...             391   \n",
       "1785   Kembalinya Tim Nasional Indonesia senior ke m...             592   \n",
       "8633   Reporter: Parto SasmitoblokBojonegoro.com - S...             153   \n",
       "8625   Perusahaan BUMN konstruksi, PT Adhi Karya Tbk...             220   \n",
       "4906   Ekonom Universitas Indonesia (UI) Faisal Basr...             342   \n",
       "\n",
       "      labels                                             corpus  \n",
       "3493       3  nyanyi muda isyana sarasvati resmi luncur albu...  \n",
       "1785       7  tim nasional indonesia senior medan laga kecew...  \n",
       "8633      20  reporter parto sasmitoblokbojonegoro com senam...  \n",
       "8625       1  usaha bumn konstruksi pt adhi karya tbk inside...  \n",
       "4906       1  ekonom universitas indonesia ui faisal basri d...  "
      ]
     },
     "execution_count": 237,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.sample(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 238,
   "id": "e28b7756",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-11-25T08:02:37.789353Z",
     "start_time": "2021-11-25T08:02:37.785376Z"
    }
   },
   "outputs": [],
   "source": [
    "def print_corpus(index):\n",
    "    example = df[df.index == index][['corpus', 'article_topic']].values[0]\n",
    "    if len(example) > 0:\n",
    "        print(example[0])\n",
    "        print('article_topic:', example[1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 239,
   "id": "79a541e9",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-11-25T08:02:37.804059Z",
     "start_time": "2021-11-25T08:02:37.795951Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "nyata spesial tagar hashtag agustus nyata tanda pasca cipta tagar media sosial twitter raya ulang tagar twitter hadir emoji khusus etik hashtag platform media sosial twitter platform media sosial mempopulerkan tagar tagar gabung kicau topik klik tagar twitter rilis daftar tagar populer indonesia penasaran daftar owibutetbisaemas salam jari kamitidaktakut savekpk euro persibday mufc melawanasap demimetallica jumatberkah deret tagar populer lihat animo masyarakat sebar topik tagar kena politik olahraga konser metallica jakarta hingga buah nyata kamitidaktakut gema pasca serang teroris tanah air tagar favorit\n",
      "article_topic: Teknologi\n"
     ]
    }
   ],
   "source": [
    "print_corpus(2)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ded86030",
   "metadata": {},
   "source": [
    "### Simple test untuk mencoba dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 240,
   "id": "e233be18",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-11-25T08:02:37.819539Z",
     "start_time": "2021-11-25T08:02:37.808116Z"
    }
   },
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.feature_extraction.text import CountVectorizer\n",
    "from sklearn.feature_extraction.text import TfidfTransformer\n",
    "from sklearn.svm import LinearSVC"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 241,
   "id": "0c7655b6",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-11-25T08:02:49.994422Z",
     "start_time": "2021-11-25T08:02:37.826298Z"
    }
   },
   "outputs": [],
   "source": [
    "X_train, X_test, y_train, y_test = train_test_split(df['corpus'], df['article_topic'], random_state = 0)\n",
    "count_vect = CountVectorizer(ngram_range=(1,2))\n",
    "X_train_counts = count_vect.fit_transform(X_train)\n",
    "tfidf_transformer = TfidfTransformer()\n",
    "X_train_tfidf = tfidf_transformer.fit_transform(X_train_counts)\n",
    "\n",
    "clf = LinearSVC().fit(X_train_tfidf, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 242,
   "id": "294b3647",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-11-25T08:02:50.021517Z",
     "start_time": "2021-11-25T08:02:49.996507Z"
    }
   },
   "outputs": [
    {
     "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>acara</th>\n",
       "      <th>adik</th>\n",
       "      <th>adil</th>\n",
       "      <th>agama</th>\n",
       "      <th>agustus</th>\n",
       "      <th>ahli</th>\n",
       "      <th>air</th>\n",
       "      <th>airlines</th>\n",
       "      <th>akibat</th>\n",
       "      <th>akomodasi</th>\n",
       "      <th>...</th>\n",
       "      <th>wajah</th>\n",
       "      <th>wanita</th>\n",
       "      <th>warga</th>\n",
       "      <th>watch</th>\n",
       "      <th>we</th>\n",
       "      <th>wilayah</th>\n",
       "      <th>wrong</th>\n",
       "      <th>yogyakarta</th>\n",
       "      <th>york</th>\n",
       "      <th>yos</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>doc1</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>doc2</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>doc3</th>\n",
       "      <td>0</td>\n",
       "      <td>6</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>doc4</th>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>4 rows × 531 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      acara  adik  adil  agama  agustus  ahli  air  airlines  akibat  \\\n",
       "doc1      0     0     1      0        1     0    0         0       1   \n",
       "doc2      0     0     0      0        0     0    0         1       0   \n",
       "doc3      0     6     0      4        0     0    0         0       0   \n",
       "doc4      2     0     0      1        0     2    1         0       0   \n",
       "\n",
       "      akomodasi  ...  wajah  wanita  warga  watch  we  wilayah  wrong  \\\n",
       "doc1          0  ...      0       0      2      1   0        1      0   \n",
       "doc2          1  ...      0       0      0      0   0        0      0   \n",
       "doc3          0  ...      1       1      0      0   1        0      0   \n",
       "doc4          0  ...      0       0      0      0   0        0      1   \n",
       "\n",
       "      yogyakarta  york  yos  \n",
       "doc1           0     3    0  \n",
       "doc2           0     0    0  \n",
       "doc3           0     0    0  \n",
       "doc4           4     0    1  \n",
       "\n",
       "[4 rows x 531 columns]"
      ]
     },
     "execution_count": 242,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "count_vect_ppt = CountVectorizer()\n",
    "for_ppt = count_vect_ppt.fit_transform(X_train[0:4])\n",
    "pd.DataFrame(for_ppt.toarray(), index=['doc1', 'doc2', 'doc3', 'doc4'], columns=count_vect_ppt.get_feature_names())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 243,
   "id": "ff0b6a06",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-11-25T08:02:50.048782Z",
     "start_time": "2021-11-25T08:02:50.023629Z"
    }
   },
   "outputs": [
    {
     "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>acara</th>\n",
       "      <th>adik</th>\n",
       "      <th>adil</th>\n",
       "      <th>agama</th>\n",
       "      <th>agustus</th>\n",
       "      <th>ahli</th>\n",
       "      <th>air</th>\n",
       "      <th>airlines</th>\n",
       "      <th>akibat</th>\n",
       "      <th>akomodasi</th>\n",
       "      <th>...</th>\n",
       "      <th>wajah</th>\n",
       "      <th>wanita</th>\n",
       "      <th>warga</th>\n",
       "      <th>watch</th>\n",
       "      <th>we</th>\n",
       "      <th>wilayah</th>\n",
       "      <th>wrong</th>\n",
       "      <th>yogyakarta</th>\n",
       "      <th>york</th>\n",
       "      <th>yos</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>doc1</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.026208</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.026208</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.026208</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.052415</td>\n",
       "      <td>0.026208</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.026208</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.078623</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>doc2</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",
       "      <td>0.048061</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.048061</td>\n",
       "      <td>...</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",
       "      <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>doc3</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.123571</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.064950</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",
       "      <td>...</td>\n",
       "      <td>0.020595</td>\n",
       "      <td>0.020595</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.020595</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>doc4</th>\n",
       "      <td>0.090036</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.035493</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.090036</td>\n",
       "      <td>0.045018</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>...</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",
       "      <td>0.045018</td>\n",
       "      <td>0.180071</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.045018</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>4 rows × 531 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "         acara      adik      adil     agama   agustus      ahli       air  \\\n",
       "doc1  0.000000  0.000000  0.026208  0.000000  0.026208  0.000000  0.000000   \n",
       "doc2  0.000000  0.000000  0.000000  0.000000  0.000000  0.000000  0.000000   \n",
       "doc3  0.000000  0.123571  0.000000  0.064950  0.000000  0.000000  0.000000   \n",
       "doc4  0.090036  0.000000  0.000000  0.035493  0.000000  0.090036  0.045018   \n",
       "\n",
       "      airlines    akibat  akomodasi  ...     wajah    wanita     warga  \\\n",
       "doc1  0.000000  0.026208   0.000000  ...  0.000000  0.000000  0.052415   \n",
       "doc2  0.048061  0.000000   0.048061  ...  0.000000  0.000000  0.000000   \n",
       "doc3  0.000000  0.000000   0.000000  ...  0.020595  0.020595  0.000000   \n",
       "doc4  0.000000  0.000000   0.000000  ...  0.000000  0.000000  0.000000   \n",
       "\n",
       "         watch        we   wilayah     wrong  yogyakarta      york       yos  \n",
       "doc1  0.026208  0.000000  0.026208  0.000000    0.000000  0.078623  0.000000  \n",
       "doc2  0.000000  0.000000  0.000000  0.000000    0.000000  0.000000  0.000000  \n",
       "doc3  0.000000  0.020595  0.000000  0.000000    0.000000  0.000000  0.000000  \n",
       "doc4  0.000000  0.000000  0.000000  0.045018    0.180071  0.000000  0.045018  \n",
       "\n",
       "[4 rows x 531 columns]"
      ]
     },
     "execution_count": 243,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tfidf_transformer_ppt = TfidfTransformer()\n",
    "td_idf_ppt = tfidf_transformer.fit_transform(for_ppt)\n",
    "pd.DataFrame(td_idf_ppt.toarray(), index=['doc1', 'doc2', 'doc3', 'doc4'], columns=count_vect_ppt.get_feature_names())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 244,
   "id": "72e4135c",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-11-25T08:02:50.054237Z",
     "start_time": "2021-11-25T08:02:50.051267Z"
    }
   },
   "outputs": [],
   "source": [
    "test_case = 'Kepolisian Inggris tengah memburu pelaku yang meledakkan bom di stasiun London. Sedikitnya 22 orang terluka dalam peristiwa yang terjadi pada Jumat pagi yang sibuk (15/9).Menurut asisten komisaris kepolisian Inggris Mark Rowley seperti dikutip The Guardian, ratusan detektif dikerahkan untuk menyelidiki kasus ini. Mereka akan memeriksa semua rekaman CCTV di jaringan kereta bawah tanah London untuk mencari pelakunya.Sebelumnya kepolisian telah menangani kasus ini sebagai aksi terorisme. Penyelidikan juga dilakukan oleh Komando Pemberantasan Terorisme Kepolisian London dan badan keamanan dalam negeri Inggris MI5.Ledakan terjadi di stasiun Parsons Green di salam satu gerbong kereta. Peristiwa ini memicu kepanikan ratusan penumpang yang langsung berhamburan menyelamatkan diri. Sedikitnya 22 orang terluka, namun tidak mengancam nyawa.Menurut kepolisian, bom rakitan itu belum sepenuhnya meledak. Tim forensik akan menyelidiki bom itu, mencari tahu bahan kimia yang digunakan, dan dari mana pelaku mempelajari pembuatannya.Inggris telah mengalami empat kali serangan teroris tahun ini, menewaskan 36 orang. Tahun 2005, 52 orang tewas dalam ledakan bunuh diri di tiga kereta bawah tanah London dan sebuah bus.'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 245,
   "id": "5ac4a9cb",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-11-25T08:02:50.061458Z",
     "start_time": "2021-11-25T08:02:50.055939Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['Internasional']\n"
     ]
    }
   ],
   "source": [
    "print(clf.predict(count_vect.transform([test_case])))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f9c2bdf2",
   "metadata": {},
   "source": [
    "## Test model begin"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b29a8391",
   "metadata": {},
   "source": [
    "I will try to use the simplest model instead of complex algorithms because my assumption is this model will be used daily, and it is easy to maintain and update the model in the future to get new data.\n",
    "\n",
    "The ML algorithm that I use for benchmarking this time is Random Forest, Naive Bayes (Multinominal), SVM Classification, and Decision Tree.\n",
    "\n",
    "I will use accuracy and choose the algorithm with the highest accuracy (with minimum threshold > 75%)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 246,
   "id": "34f75506",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-11-25T08:02:58.224564Z",
     "start_time": "2021-11-25T08:02:50.063661Z"
    }
   },
   "outputs": [],
   "source": [
    "from sklearn.feature_extraction.text import CountVectorizer\n",
    "from sklearn.feature_extraction.text import TfidfTransformer\n",
    "\n",
    "cv = CountVectorizer(ngram_range=(1,2))\n",
    "cvf = cv.fit_transform(df['corpus'].values)\n",
    "\n",
    "tfidfconverter = TfidfTransformer()\n",
    "\n",
    "features = tfidfconverter.fit_transform(cvf)\n",
    "labels = df.labels"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 247,
   "id": "7b650490",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-11-25T08:08:21.152152Z",
     "start_time": "2021-11-25T08:02:58.227271Z"
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=1)]: Using backend SequentialBackend with 1 concurrent workers.\n",
      "[Parallel(n_jobs=1)]: Done 200 out of 200 | elapsed:    6.7s finished\n",
      "[Parallel(n_jobs=1)]: Using backend SequentialBackend with 1 concurrent workers.\n",
      "[Parallel(n_jobs=1)]: Done 200 out of 200 | elapsed:    1.3s finished\n",
      "[Parallel(n_jobs=1)]: Using backend SequentialBackend with 1 concurrent workers.\n",
      "[Parallel(n_jobs=1)]: Done 200 out of 200 | elapsed:    6.5s finished\n",
      "[Parallel(n_jobs=1)]: Using backend SequentialBackend with 1 concurrent workers.\n",
      "[Parallel(n_jobs=1)]: Done 200 out of 200 | elapsed:    1.0s finished\n",
      "[Parallel(n_jobs=1)]: Using backend SequentialBackend with 1 concurrent workers.\n",
      "[Parallel(n_jobs=1)]: Done 200 out of 200 | elapsed:    6.4s finished\n",
      "[Parallel(n_jobs=1)]: Using backend SequentialBackend with 1 concurrent workers.\n",
      "[Parallel(n_jobs=1)]: Done 200 out of 200 | elapsed:    1.0s finished\n",
      "[Parallel(n_jobs=1)]: Using backend SequentialBackend with 1 concurrent workers.\n",
      "[Parallel(n_jobs=1)]: Done 200 out of 200 | elapsed:    6.4s finished\n",
      "[Parallel(n_jobs=1)]: Using backend SequentialBackend with 1 concurrent workers.\n",
      "[Parallel(n_jobs=1)]: Done 200 out of 200 | elapsed:    1.1s finished\n",
      "[Parallel(n_jobs=1)]: Using backend SequentialBackend with 1 concurrent workers.\n",
      "[Parallel(n_jobs=1)]: Done 200 out of 200 | elapsed:    6.4s finished\n",
      "[Parallel(n_jobs=1)]: Using backend SequentialBackend with 1 concurrent workers.\n",
      "[Parallel(n_jobs=1)]: Done 200 out of 200 | elapsed:    1.1s finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[LibLinear][LibLinear][LibLinear][LibLinear][LibLinear]"
     ]
    }
   ],
   "source": [
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.naive_bayes import MultinomialNB\n",
    "from sklearn.svm import LinearSVC\n",
    "from sklearn.tree import DecisionTreeClassifier\n",
    "\n",
    "from sklearn.model_selection import cross_val_score\n",
    "\n",
    "\n",
    "models = [\n",
    "    RandomForestClassifier(n_estimators=200, max_depth=3, random_state=0, criterion='entropy', verbose=1),\n",
    "    LinearSVC(verbose=1),\n",
    "    MultinomialNB(),\n",
    "    DecisionTreeClassifier(criterion = 'entropy', random_state = 0)\n",
    "]\n",
    "CV = 5\n",
    "cv_df = pd.DataFrame(index=range(CV * len(models)))\n",
    "entries = []\n",
    "for model in models:\n",
    "    model_name = model.__class__.__name__\n",
    "    accuracies = cross_val_score(model, features, labels, scoring='accuracy', cv=CV)\n",
    "    for fold_idx, accuracy in enumerate(accuracies):\n",
    "        entries.append((model_name, fold_idx, accuracy))\n",
    "cv_df = pd.DataFrame(entries, columns=['model_name', 'fold_idx', 'accuracy'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 248,
   "id": "12228b5c",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-11-25T08:08:21.162722Z",
     "start_time": "2021-11-25T08:08:21.154601Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "model_name\n",
       "DecisionTreeClassifier    0.693980\n",
       "LinearSVC                 0.879606\n",
       "MultinomialNB             0.655930\n",
       "RandomForestClassifier    0.427725\n",
       "Name: accuracy, dtype: float64"
      ]
     },
     "execution_count": 248,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cv_df.groupby('model_name').accuracy.mean()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1e00eeec",
   "metadata": {},
   "source": [
    "Karena algoritma LiniearSVC bekerja dengan baik pada dataset ini dengan akurasi 88%, saya akan menggunakannya sebagai baseline model untuk interface nanti"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 249,
   "id": "f558729f",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-11-25T08:08:26.777292Z",
     "start_time": "2021-11-25T08:08:21.165004Z"
    }
   },
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "X_train, X_test, y_train, y_test, indices_train, indices_test = train_test_split(features, labels, df.index, test_size=0.33, random_state=0)\n",
    "model = LinearSVC()\n",
    "model.fit(X_train, y_train)\n",
    "y_pred = model.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 250,
   "id": "d20a3a4c",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-11-25T08:08:31.158488Z",
     "start_time": "2021-11-25T08:08:26.779692Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x720 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.metrics import classification_report, confusion_matrix, accuracy_score\n",
    "\n",
    "conf_mat = confusion_matrix(y_test, y_pred)\n",
    "fig, ax = plt.subplots(figsize=(15,10))\n",
    "sns.heatmap(conf_mat, annot=True, fmt='d', cmap='YlGnBu',\n",
    "            xticklabels=category_id_df.article_topic.values, yticklabels=category_id_df.article_topic.values)\n",
    "plt.ylabel('Actual')\n",
    "plt.xlabel('Predicted')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 251,
   "id": "52042814",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-11-25T08:08:31.176362Z",
     "start_time": "2021-11-25T08:08:31.161126Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.83      0.88      0.86       235\n",
      "           1       0.89      0.98      0.93       594\n",
      "           2       0.94      0.85      0.90       196\n",
      "           3       0.88      0.98      0.93       455\n",
      "           4       0.99      1.00      0.99       511\n",
      "           5       0.70      0.64      0.67        25\n",
      "           6       0.80      0.44      0.57        18\n",
      "           7       0.84      0.98      0.90       380\n",
      "           8       0.70      0.33      0.45        42\n",
      "           9       0.89      0.45      0.60       146\n",
      "          10       0.86      0.44      0.59        27\n",
      "          11       0.97      0.88      0.92        75\n",
      "          12       1.00      0.83      0.91         6\n",
      "          13       0.82      0.85      0.84       186\n",
      "          14       0.00      0.00      0.00         2\n",
      "          15       0.91      0.42      0.57        24\n",
      "          16       1.00      0.21      0.35        14\n",
      "          17       0.78      0.74      0.76        19\n",
      "          18       0.85      0.65      0.74        26\n",
      "          19       0.67      0.42      0.52        19\n",
      "          20       0.95      0.92      0.93       101\n",
      "          21       0.67      0.78      0.72        65\n",
      "          22       1.00      0.62      0.77         8\n",
      "          23       1.00      0.50      0.67         4\n",
      "          24       0.78      0.70      0.74        10\n",
      "          25       0.94      0.81      0.87        57\n",
      "          26       0.00      0.00      0.00         2\n",
      "          27       0.58      1.00      0.74         7\n",
      "          28       1.00      0.60      0.75        25\n",
      "\n",
      "    accuracy                           0.89      3279\n",
      "   macro avg       0.80      0.65      0.70      3279\n",
      "weighted avg       0.89      0.89      0.88      3279\n",
      "\n",
      "0.8853308935651113\n"
     ]
    }
   ],
   "source": [
    "print(classification_report(y_test,y_pred))\n",
    "print(accuracy_score(y_test, y_pred))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "db57af6f",
   "metadata": {},
   "source": [
    "From the report, the class that is difficult to predict is the class `[5, 6, 8, 9, 10, 14, 15, 16, 19, 26]` with F-1 SCORE (Due to data imbalance) and\n",
    "To better understand the model, I would like to see which targets the model often mispredicts and how the content of articles behaves"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 267,
   "id": "15537933",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-11-25T10:34:13.248193Z",
     "start_time": "2021-11-25T10:34:13.238167Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Travel\n",
      "Personal\n",
      "Health\n",
      "Sports\n",
      "Politik\n",
      "Keuangan\n",
      "Sejarah\n",
      "Regional\n",
      "Obat-obatan\n",
      "Jakarta\n"
     ]
    }
   ],
   "source": [
    "lower_f1score_class = [5, 6, 8, 9, 10, 14, 15, 16, 19, 26]\n",
    "for n in lower_f1score_class:\n",
    "    print(id_to_category[n])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 253,
   "id": "7a9c513c",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-11-25T08:08:31.289675Z",
     "start_time": "2021-11-25T08:08:31.183387Z"
    },
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "'Ekonomi' predicted as 'Internasional' : 6 examples.\n"
     ]
    },
    {
     "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_topic</th>\n",
       "      <th>corpus</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>6168</th>\n",
       "      <td>Ekonomi</td>\n",
       "      <td>presiden amerika serikat donald trump temu ket...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5973</th>\n",
       "      <td>Ekonomi</td>\n",
       "      <td>konflik rakhine myanmar pekan sorot dunia nega...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7177</th>\n",
       "      <td>Ekonomi</td>\n",
       "      <td>toyota motor corp henti produksi pabrik rakit ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1255</th>\n",
       "      <td>Ekonomi</td>\n",
       "      <td>pt equity world futures pusat duta besar cende...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4347</th>\n",
       "      <td>Ekonomi</td>\n",
       "      <td>perintah incar wisatawan asing kunjung indones...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>326</th>\n",
       "      <td>Ekonomi</td>\n",
       "      <td>pt equityworld futures pusat harga emas turun ...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     article_topic                                             corpus\n",
       "6168       Ekonomi  presiden amerika serikat donald trump temu ket...\n",
       "5973       Ekonomi  konflik rakhine myanmar pekan sorot dunia nega...\n",
       "7177       Ekonomi  toyota motor corp henti produksi pabrik rakit ...\n",
       "1255       Ekonomi  pt equity world futures pusat duta besar cende...\n",
       "4347       Ekonomi  perintah incar wisatawan asing kunjung indones...\n",
       "326        Ekonomi  pt equityworld futures pusat harga emas turun ..."
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "'Internasional' predicted as 'Ekonomi' : 11 examples.\n"
     ]
    },
    {
     "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_topic</th>\n",
       "      <th>corpus</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>3932</th>\n",
       "      <td>Internasional</td>\n",
       "      <td>buah kapal tanker indonesia tabrak kapal keruk...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3449</th>\n",
       "      <td>Internasional</td>\n",
       "      <td>presiden indonesia joko widodo perdana menteri...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>574</th>\n",
       "      <td>Internasional</td>\n",
       "      <td>pakistan batubara indonesia juta jamshoro powe...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>994</th>\n",
       "      <td>Internasional</td>\n",
       "      <td>kjri chicago bekerjasama konsulat jenderal fil...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7961</th>\n",
       "      <td>Internasional</td>\n",
       "      <td>menteri negeri retno marsudi laku temu bilater...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2376</th>\n",
       "      <td>Internasional</td>\n",
       "      <td>usaha khadija edible oil refinery rupa impor m...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5276</th>\n",
       "      <td>Internasional</td>\n",
       "      <td>malaysia bakar perahu nelayan asing cari ikan ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1887</th>\n",
       "      <td>Internasional</td>\n",
       "      <td>panas global ancam lingkung picu tingkat angka...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4852</th>\n",
       "      <td>Internasional</td>\n",
       "      <td>perintah indonesia rencana kunjung sekretaris ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5931</th>\n",
       "      <td>Internasional</td>\n",
       "      <td>rupa bisnis investasi bangladesh hubung bilate...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4528</th>\n",
       "      <td>Internasional</td>\n",
       "      <td>tki korea selatan ketiban rejeki usaha sambal ...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      article_topic                                             corpus\n",
       "3932  Internasional  buah kapal tanker indonesia tabrak kapal keruk...\n",
       "3449  Internasional  presiden indonesia joko widodo perdana menteri...\n",
       "574   Internasional  pakistan batubara indonesia juta jamshoro powe...\n",
       "994   Internasional  kjri chicago bekerjasama konsulat jenderal fil...\n",
       "7961  Internasional  menteri negeri retno marsudi laku temu bilater...\n",
       "2376  Internasional  usaha khadija edible oil refinery rupa impor m...\n",
       "5276  Internasional  malaysia bakar perahu nelayan asing cari ikan ...\n",
       "1887  Internasional  panas global ancam lingkung picu tingkat angka...\n",
       "4852  Internasional  perintah indonesia rencana kunjung sekretaris ...\n",
       "5931  Internasional  rupa bisnis investasi bangladesh hubung bilate...\n",
       "4528  Internasional  tki korea selatan ketiban rejeki usaha sambal ..."
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "'Teknologi' predicted as 'Ekonomi' : 14 examples.\n"
     ]
    },
    {
     "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_topic</th>\n",
       "      <th>corpus</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>6003</th>\n",
       "      <td>Teknologi</td>\n",
       "      <td>podcast stuff you should know foto stuff you s...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3043</th>\n",
       "      <td>Teknologi</td>\n",
       "      <td>era serba digital situs e commerce bayar digit...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8520</th>\n",
       "      <td>Teknologi</td>\n",
       "      <td>hyperloop kembar kejut rencana hyperloop uji c...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2831</th>\n",
       "      <td>Teknologi</td>\n",
       "      <td>google kenal kreatif lingkung kantor unik appl...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3437</th>\n",
       "      <td>Teknologi</td>\n",
       "      <td>mahkamah agung ma perintah menteri hubung cabu...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39</th>\n",
       "      <td>Teknologi</td>\n",
       "      <td>ilustrasi bitcoin ethereum foto flickr berita ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>Teknologi</td>\n",
       "      <td>jalan kereta commuter line pagi ganggu anjlok ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5804</th>\n",
       "      <td>Teknologi</td>\n",
       "      <td>tahan siber indonesia pengaruh faktor utama in...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4527</th>\n",
       "      <td>Teknologi</td>\n",
       "      <td>ribut jalar dunia trading digital currency har...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8362</th>\n",
       "      <td>Teknologi</td>\n",
       "      <td>malang kabarpas com era laku ukm milik peran a...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7667</th>\n",
       "      <td>Teknologi</td>\n",
       "      <td>perintah rusia muka rencana cipta cryptocurren...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2111</th>\n",
       "      <td>Teknologi</td>\n",
       "      <td>anies baswedan sandiaga uno resmi lantik guber...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5365</th>\n",
       "      <td>Teknologi</td>\n",
       "      <td>iring kembang teknologi mudah akses transporta...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2716</th>\n",
       "      <td>Teknologi</td>\n",
       "      <td>teliti kembang hadap pesawat awak lokal tipe m...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     article_topic                                             corpus\n",
       "6003     Teknologi  podcast stuff you should know foto stuff you s...\n",
       "3043     Teknologi  era serba digital situs e commerce bayar digit...\n",
       "8520     Teknologi  hyperloop kembar kejut rencana hyperloop uji c...\n",
       "2831     Teknologi  google kenal kreatif lingkung kantor unik appl...\n",
       "3437     Teknologi  mahkamah agung ma perintah menteri hubung cabu...\n",
       "39       Teknologi  ilustrasi bitcoin ethereum foto flickr berita ...\n",
       "18       Teknologi  jalan kereta commuter line pagi ganggu anjlok ...\n",
       "5804     Teknologi  tahan siber indonesia pengaruh faktor utama in...\n",
       "4527     Teknologi  ribut jalar dunia trading digital currency har...\n",
       "8362     Teknologi  malang kabarpas com era laku ukm milik peran a...\n",
       "7667     Teknologi  perintah rusia muka rencana cipta cryptocurren...\n",
       "2111     Teknologi  anies baswedan sandiaga uno resmi lantik guber...\n",
       "5365     Teknologi  iring kembang teknologi mudah akses transporta...\n",
       "2716     Teknologi  teliti kembang hadap pesawat awak lokal tipe m..."
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "'Lifestyle' predicted as 'Ekonomi' : 6 examples.\n"
     ]
    },
    {
     "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_topic</th>\n",
       "      <th>corpus</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>9274</th>\n",
       "      <td>Lifestyle</td>\n",
       "      <td>isi parsel lihat rupa snack minum ringan baran...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9333</th>\n",
       "      <td>Lifestyle</td>\n",
       "      <td>sedap buron rumah tangga pekan garam pasar sul...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1819</th>\n",
       "      <td>Lifestyle</td>\n",
       "      <td>p t d duduk desa sumatera utara bal timur kali...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8985</th>\n",
       "      <td>Lifestyle</td>\n",
       "      <td>buah transportasi laku atur larang tumpang baw...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6883</th>\n",
       "      <td>Lifestyle</td>\n",
       "      <td>rencana wisata kuliner list restoran pilih kun...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3515</th>\n",
       "      <td>Lifestyle</td>\n",
       "      <td>besar pasak tiang nampaknya pribahasa gambar t...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     article_topic                                             corpus\n",
       "9274     Lifestyle  isi parsel lihat rupa snack minum ringan baran...\n",
       "9333     Lifestyle  sedap buron rumah tangga pekan garam pasar sul...\n",
       "1819     Lifestyle  p t d duduk desa sumatera utara bal timur kali...\n",
       "8985     Lifestyle  buah transportasi laku atur larang tumpang baw...\n",
       "6883     Lifestyle  rencana wisata kuliner list restoran pilih kun...\n",
       "3515     Lifestyle  besar pasak tiang nampaknya pribahasa gambar t..."
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "'Regional' predicted as 'Ekonomi' : 6 examples.\n"
     ]
    },
    {
     "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_topic</th>\n",
       "      <th>corpus</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>7077</th>\n",
       "      <td>Regional</td>\n",
       "      <td>kota mojokerto bangsaonline com pemkot mojoker...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7464</th>\n",
       "      <td>Regional</td>\n",
       "      <td>investasi kabupaten pasuruan catat alami tren ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4674</th>\n",
       "      <td>Regional</td>\n",
       "      <td>kalang wakil rakyat program dinas komunikasi i...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7440</th>\n",
       "      <td>Regional</td>\n",
       "      <td>temu warga dusun kedungwaru kidul desa sidepan...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2958</th>\n",
       "      <td>Regional</td>\n",
       "      <td>pacitan bangsaonline com kabupaten kota lapor ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>202</th>\n",
       "      <td>Regional</td>\n",
       "      <td>pacitan bangsaonline com marak buru ikan ilega...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     article_topic                                             corpus\n",
       "7077      Regional  kota mojokerto bangsaonline com pemkot mojoker...\n",
       "7464      Regional  investasi kabupaten pasuruan catat alami tren ...\n",
       "4674      Regional  kalang wakil rakyat program dinas komunikasi i...\n",
       "7440      Regional  temu warga dusun kedungwaru kidul desa sidepan...\n",
       "2958      Regional  pacitan bangsaonline com kabupaten kota lapor ...\n",
       "202       Regional  pacitan bangsaonline com marak buru ikan ilega..."
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "'Bojonegoro' predicted as 'Ekonomi' : 6 examples.\n"
     ]
    },
    {
     "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_topic</th>\n",
       "      <th>corpus</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>4022</th>\n",
       "      <td>Bojonegoro</td>\n",
       "      <td>reporter maratus shofifahblokbojonegoro com pe...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5996</th>\n",
       "      <td>Bojonegoro</td>\n",
       "      <td>reporter maratus shofifahblokbojonegoro com kr...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3315</th>\n",
       "      <td>Bojonegoro</td>\n",
       "      <td>reporter m yazidblokbojonegoro com letak batu ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3353</th>\n",
       "      <td>Bojonegoro</td>\n",
       "      <td>reporter m safuanblokbojonegoro com giat opera...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5248</th>\n",
       "      <td>Bojonegoro</td>\n",
       "      <td>kusnan langgan pesan sapi bupati gubernur hing...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7225</th>\n",
       "      <td>Bojonegoro</td>\n",
       "      <td>reporter nur muharromblokbojonegoro com proyek...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     article_topic                                             corpus\n",
       "4022    Bojonegoro  reporter maratus shofifahblokbojonegoro com pe...\n",
       "5996    Bojonegoro  reporter maratus shofifahblokbojonegoro com kr...\n",
       "3315    Bojonegoro  reporter m yazidblokbojonegoro com letak batu ...\n",
       "3353    Bojonegoro  reporter m safuanblokbojonegoro com giat opera...\n",
       "5248    Bojonegoro  kusnan langgan pesan sapi bupati gubernur hing...\n",
       "7225    Bojonegoro  reporter nur muharromblokbojonegoro com proyek..."
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "'Internasional' predicted as 'Hiburan' : 6 examples.\n"
     ]
    },
    {
     "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_topic</th>\n",
       "      <th>corpus</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>6065</th>\n",
       "      <td>Internasional</td>\n",
       "      <td>istana alhambra nama qalat al hamra simbol pun...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7084</th>\n",
       "      <td>Internasional</td>\n",
       "      <td>orang tenaga kerja wanita indonesia vonis adil...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6311</th>\n",
       "      <td>Internasional</td>\n",
       "      <td>harvey weinstein mundur dewan the weinstein co...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>274</th>\n",
       "      <td>Internasional</td>\n",
       "      <td>personel band death metal polandia decapitated...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>83</th>\n",
       "      <td>Internasional</td>\n",
       "      <td>indonesia gerak dab tanggap pakai agus yudhoyo...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2586</th>\n",
       "      <td>Internasional</td>\n",
       "      <td>orang tua orang anak ayah besar orang ayah chi...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      article_topic                                             corpus\n",
       "6065  Internasional  istana alhambra nama qalat al hamra simbol pun...\n",
       "7084  Internasional  orang tenaga kerja wanita indonesia vonis adil...\n",
       "6311  Internasional  harvey weinstein mundur dewan the weinstein co...\n",
       "274   Internasional  personel band death metal polandia decapitated...\n",
       "83    Internasional  indonesia gerak dab tanggap pakai agus yudhoyo...\n",
       "2586  Internasional  orang tua orang anak ayah besar orang ayah chi..."
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "'Teknologi' predicted as 'Hiburan' : 6 examples.\n"
     ]
    },
    {
     "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_topic</th>\n",
       "      <th>corpus</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>247</th>\n",
       "      <td>Teknologi</td>\n",
       "      <td>nikah momen sakral pasang suami istri langsung...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6551</th>\n",
       "      <td>Teknologi</td>\n",
       "      <td>usaha televisi gabung viva group antv tv one r...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9027</th>\n",
       "      <td>Teknologi</td>\n",
       "      <td>seri film star wars luncur bioskop desember fi...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3505</th>\n",
       "      <td>Teknologi</td>\n",
       "      <td>ajang kumpul youtuber nama tanah air internasi...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>311</th>\n",
       "      <td>Teknologi</td>\n",
       "      <td>tora sudiro mieke amalia tampil bintang iklan ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3833</th>\n",
       "      <td>Teknologi</td>\n",
       "      <td>konten tarik akun sosial media perkara mudah k...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     article_topic                                             corpus\n",
       "247      Teknologi  nikah momen sakral pasang suami istri langsung...\n",
       "6551     Teknologi  usaha televisi gabung viva group antv tv one r...\n",
       "9027     Teknologi  seri film star wars luncur bioskop desember fi...\n",
       "3505     Teknologi  ajang kumpul youtuber nama tanah air internasi...\n",
       "311      Teknologi  tora sudiro mieke amalia tampil bintang iklan ...\n",
       "3833     Teknologi  konten tarik akun sosial media perkara mudah k..."
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "'Lifestyle' predicted as 'Hiburan' : 11 examples.\n"
     ]
    },
    {
     "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_topic</th>\n",
       "      <th>corpus</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1194</th>\n",
       "      <td>Lifestyle</td>\n",
       "      <td>pagi berita duka dr ryan thamrin pandu acara d...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8518</th>\n",
       "      <td>Lifestyle</td>\n",
       "      <td>menteri budaya olahraga pariwisata korea selat...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1874</th>\n",
       "      <td>Lifestyle</td>\n",
       "      <td>film horor seru nontonnya ajak teman anti nont...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8528</th>\n",
       "      <td>Lifestyle</td>\n",
       "      <td>libur ngobrol mata anak perempuan remaja ngobr...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2571</th>\n",
       "      <td>Lifestyle</td>\n",
       "      <td>pro kontra suami istri alias poligami isu hang...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7254</th>\n",
       "      <td>Lifestyle</td>\n",
       "      <td>artikel bagus bahas komik donald duck coba bed...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9826</th>\n",
       "      <td>Lifestyle</td>\n",
       "      <td>pilih nama bayi gampang gampang susah indah de...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6938</th>\n",
       "      <td>Lifestyle</td>\n",
       "      <td>wingardium leviosa cinta film fantasi sihir ta...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3214</th>\n",
       "      <td>Lifestyle</td>\n",
       "      <td>aktris cantik darah sunda minangkabau jawa res...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1955</th>\n",
       "      <td>Lifestyle</td>\n",
       "      <td>sepakat hasil suami anggota keluarga dukung is...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7587</th>\n",
       "      <td>Lifestyle</td>\n",
       "      <td>cinta binatang kucing rupa salah hewan manja t...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     article_topic                                             corpus\n",
       "1194     Lifestyle  pagi berita duka dr ryan thamrin pandu acara d...\n",
       "8518     Lifestyle  menteri budaya olahraga pariwisata korea selat...\n",
       "1874     Lifestyle  film horor seru nontonnya ajak teman anti nont...\n",
       "8528     Lifestyle  libur ngobrol mata anak perempuan remaja ngobr...\n",
       "2571     Lifestyle  pro kontra suami istri alias poligami isu hang...\n",
       "7254     Lifestyle  artikel bagus bahas komik donald duck coba bed...\n",
       "9826     Lifestyle  pilih nama bayi gampang gampang susah indah de...\n",
       "6938     Lifestyle  wingardium leviosa cinta film fantasi sihir ta...\n",
       "3214     Lifestyle  aktris cantik darah sunda minangkabau jawa res...\n",
       "1955     Lifestyle  sepakat hasil suami anggota keluarga dukung is...\n",
       "7587     Lifestyle  cinta binatang kucing rupa salah hewan manja t..."
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "'K-Pop' predicted as 'Hiburan' : 10 examples.\n"
     ]
    },
    {
     "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_topic</th>\n",
       "      <th>corpus</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>8239</th>\n",
       "      <td>K-Pop</td>\n",
       "      <td>dreamcatcher kejut gemar rilis buah video inda...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1533</th>\n",
       "      <td>K-Pop</td>\n",
       "      <td>pentagon buah video dance latih koreografi lik...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6049</th>\n",
       "      <td>K-Pop</td>\n",
       "      <td>idola k pop jalin kisah asmara orang perkara m...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7493</th>\n",
       "      <td>K-Pop</td>\n",
       "      <td>pentol boyband bigbang kwon jiyoung g dragon g...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1127</th>\n",
       "      <td>K-Pop</td>\n",
       "      <td>acara countdown asi games gelar malam kawasan ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2550</th>\n",
       "      <td>K-Pop</td>\n",
       "      <td>mantan anggota girl group after school jung ah...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7055</th>\n",
       "      <td>K-Pop</td>\n",
       "      <td>rahasia korea identik k pop k drama jepang ide...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8802</th>\n",
       "      <td>K-Pop</td>\n",
       "      <td>fans exo baper member exo china kris luhan tao...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8430</th>\n",
       "      <td>K-Pop</td>\n",
       "      <td>idola keikutsertaan acara survival entertainme...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5553</th>\n",
       "      <td>K-Pop</td>\n",
       "      <td>sechskies comeback album rilis rangkai teaser ...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     article_topic                                             corpus\n",
       "8239         K-Pop  dreamcatcher kejut gemar rilis buah video inda...\n",
       "1533         K-Pop  pentagon buah video dance latih koreografi lik...\n",
       "6049         K-Pop  idola k pop jalin kisah asmara orang perkara m...\n",
       "7493         K-Pop  pentol boyband bigbang kwon jiyoung g dragon g...\n",
       "1127         K-Pop  acara countdown asi games gelar malam kawasan ...\n",
       "2550         K-Pop  mantan anggota girl group after school jung ah...\n",
       "7055         K-Pop  rahasia korea identik k pop k drama jepang ide...\n",
       "8802         K-Pop  fans exo baper member exo china kris luhan tao...\n",
       "8430         K-Pop  idola keikutsertaan acara survival entertainme...\n",
       "5553         K-Pop  sechskies comeback album rilis rangkai teaser ..."
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "'Sports' predicted as 'Sepak Bola' : 72 examples.\n"
     ]
    },
    {
     "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_topic</th>\n",
       "      <th>corpus</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>665</th>\n",
       "      <td>Sports</td>\n",
       "      <td>virus fifa main tim tim besar eropa alam ceder...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2162</th>\n",
       "      <td>Sports</td>\n",
       "      <td>momok lawas hantu arsenal francis coquelin mes...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7042</th>\n",
       "      <td>Sports</td>\n",
       "      <td>arsenal takluk norwich city skor ajang babak b...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1035</th>\n",
       "      <td>Sports</td>\n",
       "      <td>manuel neuer praktis berita absen laga lawan s...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8794</th>\n",
       "      <td>Sports</td>\n",
       "      <td>manchester city catat sempurna premier league ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3671</th>\n",
       "      <td>Sports</td>\n",
       "      <td>laga big match temu liverpool manchester unite...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1303</th>\n",
       "      <td>Sports</td>\n",
       "      <td>musim manchester united tampil buruk premier l...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1329</th>\n",
       "      <td>Sports</td>\n",
       "      <td>bursa transfer klub klub inggris tutup agustus...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8188</th>\n",
       "      <td>Sports</td>\n",
       "      <td>chelsea hasil raih poin penuh lawat bet stadiu...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1953</th>\n",
       "      <td>Sports</td>\n",
       "      <td>menang hasil buku tim nasional indonesia u aja...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>72 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     article_topic                                             corpus\n",
       "665         Sports  virus fifa main tim tim besar eropa alam ceder...\n",
       "2162        Sports  momok lawas hantu arsenal francis coquelin mes...\n",
       "7042        Sports  arsenal takluk norwich city skor ajang babak b...\n",
       "1035        Sports  manuel neuer praktis berita absen laga lawan s...\n",
       "8794        Sports  manchester city catat sempurna premier league ...\n",
       "...            ...                                                ...\n",
       "3671        Sports  laga big match temu liverpool manchester unite...\n",
       "1303        Sports  musim manchester united tampil buruk premier l...\n",
       "1329        Sports  bursa transfer klub klub inggris tutup agustus...\n",
       "8188        Sports  chelsea hasil raih poin penuh lawat bet stadiu...\n",
       "1953        Sports  menang hasil buku tim nasional indonesia u aja...\n",
       "\n",
       "[72 rows x 2 columns]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "'Internasional' predicted as 'Lifestyle' : 6 examples.\n"
     ]
    },
    {
     "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_topic</th>\n",
       "      <th>corpus</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>5381</th>\n",
       "      <td>Internasional</td>\n",
       "      <td>sampah sampah selokan indonesia daun plastik b...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4913</th>\n",
       "      <td>Internasional</td>\n",
       "      <td>perintah inggris resmi larang edar cokelat bat...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3032</th>\n",
       "      <td>Internasional</td>\n",
       "      <td>buah teliti kendali mimpi kontrol mimpi sensas...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3118</th>\n",
       "      <td>Internasional</td>\n",
       "      <td>marak potret hamil gaya gemas emily mueller pe...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9704</th>\n",
       "      <td>Internasional</td>\n",
       "      <td>canggih cepat teknologi mudah hidup manusia or...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5607</th>\n",
       "      <td>Internasional</td>\n",
       "      <td>raya takhta sultan hassanal bolkiah mu izzaddi...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      article_topic                                             corpus\n",
       "5381  Internasional  sampah sampah selokan indonesia daun plastik b...\n",
       "4913  Internasional  perintah inggris resmi larang edar cokelat bat...\n",
       "3032  Internasional  buah teliti kendali mimpi kontrol mimpi sensas...\n",
       "3118  Internasional  marak potret hamil gaya gemas emily mueller pe...\n",
       "9704  Internasional  canggih cepat teknologi mudah hidup manusia or...\n",
       "5607  Internasional  raya takhta sultan hassanal bolkiah mu izzaddi..."
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "'Health' predicted as 'Lifestyle' : 9 examples.\n"
     ]
    },
    {
     "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_topic</th>\n",
       "      <th>corpus</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>7555</th>\n",
       "      <td>Health</td>\n",
       "      <td>selamat postingan admin bahas artikel manfaat ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7307</th>\n",
       "      <td>Health</td>\n",
       "      <td>the healthy perfect food slogan sebar perintah...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5668</th>\n",
       "      <td>Health</td>\n",
       "      <td>walatra sehat mata softgel walatra sehat mata ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7987</th>\n",
       "      <td>Health</td>\n",
       "      <td>orang tua patut hati hati sakit infeksi paru p...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2580</th>\n",
       "      <td>Health</td>\n",
       "      <td>kanker lidah salah jenis kanker mulut mulut ka...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6523</th>\n",
       "      <td>Health</td>\n",
       "      <td>pasang nikah milik anak pasang nikah hitung ka...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5526</th>\n",
       "      <td>Health</td>\n",
       "      <td>teh hijau percaya jaga sehat orang orang china...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1249</th>\n",
       "      <td>Health</td>\n",
       "      <td>mangkuk nasi matang kandung kalori tak bakar u...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3326</th>\n",
       "      <td>Health</td>\n",
       "      <td>jeruk rupa salah jenis buah milik gemar buah j...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     article_topic                                             corpus\n",
       "7555        Health  selamat postingan admin bahas artikel manfaat ...\n",
       "7307        Health  the healthy perfect food slogan sebar perintah...\n",
       "5668        Health  walatra sehat mata softgel walatra sehat mata ...\n",
       "7987        Health  orang tua patut hati hati sakit infeksi paru p...\n",
       "2580        Health  kanker lidah salah jenis kanker mulut mulut ka...\n",
       "6523        Health  pasang nikah milik anak pasang nikah hitung ka...\n",
       "5526        Health  teh hijau percaya jaga sehat orang orang china...\n",
       "1249        Health  mangkuk nasi matang kandung kalori tak bakar u...\n",
       "3326        Health  jeruk rupa salah jenis buah milik gemar buah j..."
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "'Health' predicted as 'Kesehatan' : 13 examples.\n"
     ]
    },
    {
     "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_topic</th>\n",
       "      <th>corpus</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>6683</th>\n",
       "      <td>Health</td>\n",
       "      <td>selamat distory zahra herbal gejala kanker lid...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5135</th>\n",
       "      <td>Health</td>\n",
       "      <td>hallo bunda moga tegar hadap gugur ledies meng...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2760</th>\n",
       "      <td>Health</td>\n",
       "      <td>g sea jelly jelly gamat walatra g sea jelly je...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2948</th>\n",
       "      <td>Health</td>\n",
       "      <td>kolera rupa infeksi bakteri derita derita dehi...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4454</th>\n",
       "      <td>Health</td>\n",
       "      <td>bersih telinga bersih cotton bud dorong earwas...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3641</th>\n",
       "      <td>Health</td>\n",
       "      <td>sembuh asam lambung totalpenyakit asam lambung...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1892</th>\n",
       "      <td>Health</td>\n",
       "      <td>obat nyeri sendi timbul sakit orang salep gel ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8444</th>\n",
       "      <td>Health</td>\n",
       "      <td>sistem nafas manusia milik tugas sulit orang n...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9482</th>\n",
       "      <td>Health</td>\n",
       "      <td>lupus salah sakit kronis serang tubuh manusia ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5198</th>\n",
       "      <td>Health</td>\n",
       "      <td>kantuk ngantuk rupa sifat alami manusia lepas ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6491</th>\n",
       "      <td>Health</td>\n",
       "      <td>batu kandung kemih orang derita sakit batu gin...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1343</th>\n",
       "      <td>Health</td>\n",
       "      <td>besar orang kondisi fisik prima rupa buah nila...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5768</th>\n",
       "      <td>Health</td>\n",
       "      <td>nyanyi sena gomez jalan transplantasi ginjal s...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     article_topic                                             corpus\n",
       "6683        Health  selamat distory zahra herbal gejala kanker lid...\n",
       "5135        Health  hallo bunda moga tegar hadap gugur ledies meng...\n",
       "2760        Health  g sea jelly jelly gamat walatra g sea jelly je...\n",
       "2948        Health  kolera rupa infeksi bakteri derita derita dehi...\n",
       "4454        Health  bersih telinga bersih cotton bud dorong earwas...\n",
       "3641        Health  sembuh asam lambung totalpenyakit asam lambung...\n",
       "1892        Health  obat nyeri sendi timbul sakit orang salep gel ...\n",
       "8444        Health  sistem nafas manusia milik tugas sulit orang n...\n",
       "9482        Health  lupus salah sakit kronis serang tubuh manusia ...\n",
       "5198        Health  kantuk ngantuk rupa sifat alami manusia lepas ...\n",
       "6491        Health  batu kandung kemih orang derita sakit batu gin...\n",
       "1343        Health  besar orang kondisi fisik prima rupa buah nila...\n",
       "5768        Health  nyanyi sena gomez jalan transplantasi ginjal s..."
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "'Obat-obatan' predicted as 'Kesehatan' : 6 examples.\n"
     ]
    },
    {
     "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_topic</th>\n",
       "      <th>corpus</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>7495</th>\n",
       "      <td>Obat-obatan</td>\n",
       "      <td>sakit radang usus buntu radang usus buntu kond...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>722</th>\n",
       "      <td>Obat-obatan</td>\n",
       "      <td>sakit wasir istilah medis hemoroid bengkak bul...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6169</th>\n",
       "      <td>Obat-obatan</td>\n",
       "      <td>obat osteoporosis herbal alami wanita ampuh da...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6070</th>\n",
       "      <td>Obat-obatan</td>\n",
       "      <td>sakit miom sakit miom kondisi tumbuh tumor atu...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5606</th>\n",
       "      <td>Obat-obatan</td>\n",
       "      <td>epigastrium kenal nyeri ulu hati sakit akibat ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>963</th>\n",
       "      <td>Obat-obatan</td>\n",
       "      <td>ampuh hilang bekas jerawat cepat jerawat kulit...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     article_topic                                             corpus\n",
       "7495   Obat-obatan  sakit radang usus buntu radang usus buntu kond...\n",
       "722    Obat-obatan  sakit wasir istilah medis hemoroid bengkak bul...\n",
       "6169   Obat-obatan  obat osteoporosis herbal alami wanita ampuh da...\n",
       "6070   Obat-obatan  sakit miom sakit miom kondisi tumbuh tumor atu...\n",
       "5606   Obat-obatan  epigastrium kenal nyeri ulu hati sakit akibat ...\n",
       "963    Obat-obatan  ampuh hilang bekas jerawat cepat jerawat kulit..."
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "from IPython.display import display\n",
    "\n",
    "for predicted in category_id_df.labels:\n",
    "    for actual in category_id_df.labels:\n",
    "        if predicted != actual and conf_mat[actual, predicted] >= 6:\n",
    "            print(\"'{}' predicted as '{}' : {} examples.\".format(id_to_category[actual], id_to_category[predicted], conf_mat[actual, predicted]))\n",
    "            display(df.loc[indices_test[(y_test == actual) & (y_pred == predicted)]][['article_topic', 'corpus']])\n",
    "            print('')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4b12dd7f",
   "metadata": {},
   "source": [
    "Okay, in some classes, the article content does (assumes) contain words that describe a high correlation to other classes so that the model becomes incorrect in predictions, for example:\n",
    "- President of `United States Donald Trump will . . . in the prediction of 'International' when it should be 'Economics.'\n",
    "- What is the `disease` of appendicitis.? Inflammation of the intestines... predicts 'Health' when it should be 'Medicine.'\n",
    "- and some target class `health` predicted `health` which are indeed similar in terms of word correlation in content"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "efc023a1",
   "metadata": {},
   "source": [
    "# Final Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 274,
   "id": "7f2bc88f",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-11-25T11:16:33.658310Z",
     "start_time": "2021-11-25T11:16:33.653051Z"
    }
   },
   "outputs": [],
   "source": [
    "final_model = LinearSVC(verbose=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 275,
   "id": "17c9348a",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-11-25T11:16:35.900496Z",
     "start_time": "2021-11-25T11:16:33.952665Z"
    }
   },
   "outputs": [],
   "source": [
    "final_cv = CountVectorizer()\n",
    "cvf = final_cv.fit_transform(df['corpus'].values)\n",
    "final_tfidfconverter = TfidfTransformer()\n",
    "features = final_tfidfconverter.fit_transform(cvf)\n",
    "labels = df.labels"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 276,
   "id": "c90886bd",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-11-25T11:16:37.622517Z",
     "start_time": "2021-11-25T11:16:35.903361Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[LibLinear]"
     ]
    },
    {
     "data": {
      "text/plain": [
       "LinearSVC(C=1.0, class_weight=None, dual=True, fit_intercept=True,\n",
       "          intercept_scaling=1, loss='squared_hinge', max_iter=1000,\n",
       "          multi_class='ovr', penalty='l2', random_state=None, tol=0.0001,\n",
       "          verbose=1)"
      ]
     },
     "execution_count": 276,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "final_model.fit(features, labels)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 277,
   "id": "8727cb39",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-11-25T11:16:37.629274Z",
     "start_time": "2021-11-25T11:16:37.625376Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Internasional\n"
     ]
    }
   ],
   "source": [
    "prediction = final_model.predict(final_cv.transform([test_case]))\n",
    "print(id_to_category[int(prediction)])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 278,
   "id": "5b739976",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-11-25T11:16:37.645781Z",
     "start_time": "2021-11-25T11:16:37.633783Z"
    }
   },
   "outputs": [],
   "source": [
    "test = pd.read_csv(\"test_set.csv\")\n",
    "test_case2 = test['article_content'].values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 279,
   "id": "65f3d9f5",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-11-25T11:16:37.661889Z",
     "start_time": "2021-11-25T11:16:37.651888Z"
    }
   },
   "outputs": [
    {
     "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_id</th>\n",
       "      <th>article_topic</th>\n",
       "      <th>article_content</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>93205794</td>\n",
       "      <td>Internasional</td>\n",
       "      <td>Kepolisian Inggris tengah memburu pelaku yang...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   article_id  article_topic  \\\n",
       "0    93205794  Internasional   \n",
       "\n",
       "                                     article_content  \n",
       "0   Kepolisian Inggris tengah memburu pelaku yang...  "
      ]
     },
     "execution_count": 279,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 280,
   "id": "138fa638",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-11-25T11:16:37.670570Z",
     "start_time": "2021-11-25T11:16:37.664562Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Internasional\n"
     ]
    }
   ],
   "source": [
    "prediction = final_model.predict(final_cv.transform([str(test_case2)]))\n",
    "label = id_to_category[int(prediction)]\n",
    "print(label)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "66af23d4",
   "metadata": {},
   "source": [
    "# Improvement"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 287,
   "id": "4efc8c0a",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-11-27T06:12:11.708847Z",
     "start_time": "2021-11-27T06:11:18.430146Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "# 'Bisnis':\n",
      "  . Most correlated bigrams:\n",
      "       . arsa batas\n",
      "       . amsterdam jerman\n",
      "  . Most correlated unigrams:\n",
      "       . almukaromah\n",
      "       . abp\n",
      "# 'Bojonegoro':\n",
      "  . Most correlated bigrams:\n",
      "       . aff malam\n",
      "       . afganistan tuduh\n",
      "  . Most correlated unigrams:\n",
      "       . alamiin\n",
      "       . alexa\n",
      "# 'Ekonomi':\n",
      "  . Most correlated bigrams:\n",
      "       . argumen thaler\n",
      "       . ariyo jumpa\n",
      "  . Most correlated unigrams:\n",
      "       . ajuin\n",
      "       . atmoko\n",
      "# 'Haji':\n",
      "  . Most correlated bigrams:\n",
      "       . aliansi filipina\n",
      "       . alam ahli\n",
      "  . Most correlated unigrams:\n",
      "       . authentic\n",
      "       . algiers\n",
      "# 'Health':\n",
      "  . Most correlated bigrams:\n",
      "       . angkut jemaah\n",
      "       . aktifitas ruang\n",
      "  . Most correlated unigrams:\n",
      "       . academy\n",
      "       . aliano\n",
      "# 'Hiburan':\n",
      "  . Most correlated bigrams:\n",
      "       . aket hadiah\n",
      "       . akses tumpang\n",
      "  . Most correlated unigrams:\n",
      "       . abf\n",
      "       . angke\n",
      "# 'Horor':\n",
      "  . Most correlated bigrams:\n",
      "       . action hollywood\n",
      "       . andrey langsung\n",
      "  . Most correlated unigrams:\n",
      "       . awapi\n",
      "       . anterior\n",
      "# 'Hukum':\n",
      "  . Most correlated bigrams:\n",
      "       . andres maryland\n",
      "       . amatir gacy\n",
      "  . Most correlated unigrams:\n",
      "       . acontoh\n",
      "       . applikasi\n",
      "# 'Internasional':\n",
      "  . Most correlated bigrams:\n",
      "       . asuransi duga\n",
      "       . areola psg\n",
      "  . Most correlated unigrams:\n",
      "       . aqiel\n",
      "       . arianto\n",
      "# 'Jakarta':\n",
      "  . Most correlated bigrams:\n",
      "       . ajar lihat\n",
      "       . anfield road\n",
      "  . Most correlated unigrams:\n",
      "       . affu\n",
      "       . anandia\n",
      "# 'K-Pop':\n",
      "  . Most correlated bigrams:\n",
      "       . anggap sanksi\n",
      "       . agustus operasi\n",
      "  . Most correlated unigrams:\n",
      "       . aula\n",
      "       . abdisatrijo\n",
      "# 'KPK':\n",
      "  . Most correlated bigrams:\n",
      "       . asn jamaah\n",
      "       . atlet thailand\n",
      "  . Most correlated unigrams:\n",
      "       . alamiahglucosaminechondroitinprotein\n",
      "       . apart\n",
      "# 'Kesehatan':\n",
      "  . Most correlated bigrams:\n",
      "       . apresiasi positif\n",
      "       . aktifitas ruang\n",
      "  . Most correlated unigrams:\n",
      "       . abar\n",
      "       . aliano\n",
      "# 'Keuangan':\n",
      "  . Most correlated bigrams:\n",
      "       . alexander lay\n",
      "       . aktif rusak\n",
      "  . Most correlated unigrams:\n",
      "       . aksan\n",
      "       . alkilbenzena\n",
      "# 'Lifestyle':\n",
      "  . Most correlated bigrams:\n",
      "       . akrab gelar\n",
      "       . anak aksi\n",
      "  . Most correlated unigrams:\n",
      "       . arofah\n",
      "       . agentdipegang\n",
      "# 'MotoGP':\n",
      "  . Most correlated bigrams:\n",
      "       . anemia asupan\n",
      "       . ade vicky\n",
      "  . Most correlated unigrams:\n",
      "       . alive\n",
      "       . apestak\n",
      "# 'Obat-obatan':\n",
      "  . Most correlated bigrams:\n",
      "       . angkut jemaah\n",
      "       . apresiasi positif\n",
      "  . Most correlated unigrams:\n",
      "       . abar\n",
      "       . aliano\n",
      "# 'Otomotif':\n",
      "  . Most correlated bigrams:\n",
      "       . ayah kevin\n",
      "       . atlet salah\n",
      "  . Most correlated unigrams:\n",
      "       . affair\n",
      "       . adailan\n",
      "# 'Pendidikan':\n",
      "  . Most correlated bigrams:\n",
      "       . al latih\n",
      "       . as imam\n",
      "  . Most correlated unigrams:\n",
      "       . arthaputra\n",
      "       . abdurahman\n",
      "# 'Personal':\n",
      "  . Most correlated bigrams:\n",
      "       . alida kerabat\n",
      "       . alam rintis\n",
      "  . Most correlated unigrams:\n",
      "       . addendum\n",
      "       . akar\n",
      "# 'Pilgub Jatim':\n",
      "  . Most correlated bigrams:\n",
      "       . ali jalur\n",
      "       . always remember\n",
      "  . Most correlated unigrams:\n",
      "       . aby\n",
      "       . aslavia\n",
      "# 'Politik':\n",
      "  . Most correlated bigrams:\n",
      "       . apartemen paris\n",
      "       . anung herlianto\n",
      "  . Most correlated unigrams:\n",
      "       . arteriosklerosis\n",
      "       . antarpersonel\n",
      "# 'Regional':\n",
      "  . Most correlated bigrams:\n",
      "       . adkerson jonan\n",
      "       . anniesa hadap\n",
      "  . Most correlated unigrams:\n",
      "       . asdep\n",
      "       . aol\n",
      "# 'Sains':\n",
      "  . Most correlated bigrams:\n",
      "       . atur indadari\n",
      "       . arena bsd\n",
      "  . Most correlated unigrams:\n",
      "       . aristoteles\n",
      "       . aktifitasku\n",
      "# 'Sejarah':\n",
      "  . Most correlated bigrams:\n",
      "       . adinda paut\n",
      "       . afeksi gambar\n",
      "  . Most correlated unigrams:\n",
      "       . amburadul\n",
      "       . anis\n",
      "# 'Sepak Bola':\n",
      "  . Most correlated bigrams:\n",
      "       . akun laku\n",
      "       . ana keluarga\n",
      "  . Most correlated unigrams:\n",
      "       . apkikasi\n",
      "       . atak\n",
      "# 'Sports':\n",
      "  . Most correlated bigrams:\n",
      "       . acu nikkei\n",
      "       . ade vicky\n",
      "  . Most correlated unigrams:\n",
      "       . aremania\n",
      "       . aureus\n",
      "# 'Teknologi':\n",
      "  . Most correlated bigrams:\n",
      "       . aldiano pokok\n",
      "       . api menteri\n",
      "  . Most correlated unigrams:\n",
      "       . associataed\n",
      "       . armory\n",
      "# 'Travel':\n",
      "  . Most correlated bigrams:\n",
      "       . arsenal jamu\n",
      "       . aktor utama\n",
      "  . Most correlated unigrams:\n",
      "       . ali\n",
      "       . adhiguna\n"
     ]
    }
   ],
   "source": [
    "from sklearn.feature_selection import chi2\n",
    "import numpy as np\n",
    "\n",
    "N = 2\n",
    "bigram_unigram = []\n",
    "for article_topic, label in sorted(category_to_id.items()):\n",
    "    features_chi2 = chi2(features, labels == label)\n",
    "    indices = np.argsort(features_chi2[0])\n",
    "    feature_names = np.array(cv.get_feature_names())[indices]\n",
    "    \n",
    "    unigrams = [v for v in feature_names if len(v.split(' ')) == 1]\n",
    "    bigrams = [v for v in feature_names if len(v.split(' ')) == 2]\n",
    "    \n",
    "    print(\"# '{}':\".format(article_topic))\n",
    "    print(\"  . Most correlated bigrams:\\n       . {}\".format('\\n       . '.join(bigrams[-N:])))\n",
    "    print(\"  . Most correlated unigrams:\\n       . {}\".format('\\n       . '.join(unigrams[-N:])))\n",
    "    prepared_data = {\n",
    "        \"topic\":article_topic,\n",
    "        \"bigrams\":bigrams[-N:],\n",
    "        \"unigrams\": unigrams[-N:]\n",
    "    }\n",
    "    bigram_unigram.append(prepared_data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 288,
   "id": "994fcb70",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-11-27T06:12:11.746144Z",
     "start_time": "2021-11-27T06:12:11.711138Z"
    }
   },
   "outputs": [
    {
     "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>topic</th>\n",
       "      <th>bigrams</th>\n",
       "      <th>unigrams</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Bisnis</td>\n",
       "      <td>[arsa batas, amsterdam jerman]</td>\n",
       "      <td>[almukaromah, abp]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Bojonegoro</td>\n",
       "      <td>[aff malam, afganistan tuduh]</td>\n",
       "      <td>[alamiin, alexa]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Ekonomi</td>\n",
       "      <td>[argumen thaler, ariyo jumpa]</td>\n",
       "      <td>[ajuin, atmoko]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Haji</td>\n",
       "      <td>[aliansi filipina, alam ahli]</td>\n",
       "      <td>[authentic, algiers]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Health</td>\n",
       "      <td>[angkut jemaah, aktifitas ruang]</td>\n",
       "      <td>[academy, aliano]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>Hiburan</td>\n",
       "      <td>[aket hadiah, akses tumpang]</td>\n",
       "      <td>[abf, angke]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>Horor</td>\n",
       "      <td>[action hollywood, andrey langsung]</td>\n",
       "      <td>[awapi, anterior]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>Hukum</td>\n",
       "      <td>[andres maryland, amatir gacy]</td>\n",
       "      <td>[acontoh, applikasi]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>Internasional</td>\n",
       "      <td>[asuransi duga, areola psg]</td>\n",
       "      <td>[aqiel, arianto]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>Jakarta</td>\n",
       "      <td>[ajar lihat, anfield road]</td>\n",
       "      <td>[affu, anandia]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>K-Pop</td>\n",
       "      <td>[anggap sanksi, agustus operasi]</td>\n",
       "      <td>[aula, abdisatrijo]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>KPK</td>\n",
       "      <td>[asn jamaah, atlet thailand]</td>\n",
       "      <td>[alamiahglucosaminechondroitinprotein, apart]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>Kesehatan</td>\n",
       "      <td>[apresiasi positif, aktifitas ruang]</td>\n",
       "      <td>[abar, aliano]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>Keuangan</td>\n",
       "      <td>[alexander lay, aktif rusak]</td>\n",
       "      <td>[aksan, alkilbenzena]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>Lifestyle</td>\n",
       "      <td>[akrab gelar, anak aksi]</td>\n",
       "      <td>[arofah, agentdipegang]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>MotoGP</td>\n",
       "      <td>[anemia asupan, ade vicky]</td>\n",
       "      <td>[alive, apestak]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>Obat-obatan</td>\n",
       "      <td>[angkut jemaah, apresiasi positif]</td>\n",
       "      <td>[abar, aliano]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>Otomotif</td>\n",
       "      <td>[ayah kevin, atlet salah]</td>\n",
       "      <td>[affair, adailan]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>Pendidikan</td>\n",
       "      <td>[al latih, as imam]</td>\n",
       "      <td>[arthaputra, abdurahman]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>Personal</td>\n",
       "      <td>[alida kerabat, alam rintis]</td>\n",
       "      <td>[addendum, akar]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>Pilgub Jatim</td>\n",
       "      <td>[ali jalur, always remember]</td>\n",
       "      <td>[aby, aslavia]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>Politik</td>\n",
       "      <td>[apartemen paris, anung herlianto]</td>\n",
       "      <td>[arteriosklerosis, antarpersonel]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>Regional</td>\n",
       "      <td>[adkerson jonan, anniesa hadap]</td>\n",
       "      <td>[asdep, aol]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>Sains</td>\n",
       "      <td>[atur indadari, arena bsd]</td>\n",
       "      <td>[aristoteles, aktifitasku]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>Sejarah</td>\n",
       "      <td>[adinda paut, afeksi gambar]</td>\n",
       "      <td>[amburadul, anis]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>Sepak Bola</td>\n",
       "      <td>[akun laku, ana keluarga]</td>\n",
       "      <td>[apkikasi, atak]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>Sports</td>\n",
       "      <td>[acu nikkei, ade vicky]</td>\n",
       "      <td>[aremania, aureus]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>Teknologi</td>\n",
       "      <td>[aldiano pokok, api menteri]</td>\n",
       "      <td>[associataed, armory]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>Travel</td>\n",
       "      <td>[arsenal jamu, aktor utama]</td>\n",
       "      <td>[ali, adhiguna]</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            topic                               bigrams  \\\n",
       "0          Bisnis        [arsa batas, amsterdam jerman]   \n",
       "1      Bojonegoro         [aff malam, afganistan tuduh]   \n",
       "2         Ekonomi         [argumen thaler, ariyo jumpa]   \n",
       "3            Haji         [aliansi filipina, alam ahli]   \n",
       "4          Health      [angkut jemaah, aktifitas ruang]   \n",
       "5         Hiburan          [aket hadiah, akses tumpang]   \n",
       "6           Horor   [action hollywood, andrey langsung]   \n",
       "7           Hukum        [andres maryland, amatir gacy]   \n",
       "8   Internasional           [asuransi duga, areola psg]   \n",
       "9         Jakarta            [ajar lihat, anfield road]   \n",
       "10          K-Pop      [anggap sanksi, agustus operasi]   \n",
       "11            KPK          [asn jamaah, atlet thailand]   \n",
       "12      Kesehatan  [apresiasi positif, aktifitas ruang]   \n",
       "13       Keuangan          [alexander lay, aktif rusak]   \n",
       "14      Lifestyle              [akrab gelar, anak aksi]   \n",
       "15         MotoGP            [anemia asupan, ade vicky]   \n",
       "16    Obat-obatan    [angkut jemaah, apresiasi positif]   \n",
       "17       Otomotif             [ayah kevin, atlet salah]   \n",
       "18     Pendidikan                   [al latih, as imam]   \n",
       "19       Personal          [alida kerabat, alam rintis]   \n",
       "20   Pilgub Jatim          [ali jalur, always remember]   \n",
       "21        Politik    [apartemen paris, anung herlianto]   \n",
       "22       Regional       [adkerson jonan, anniesa hadap]   \n",
       "23          Sains            [atur indadari, arena bsd]   \n",
       "24        Sejarah          [adinda paut, afeksi gambar]   \n",
       "25     Sepak Bola             [akun laku, ana keluarga]   \n",
       "26         Sports               [acu nikkei, ade vicky]   \n",
       "27      Teknologi          [aldiano pokok, api menteri]   \n",
       "28         Travel           [arsenal jamu, aktor utama]   \n",
       "\n",
       "                                         unigrams  \n",
       "0                              [almukaromah, abp]  \n",
       "1                                [alamiin, alexa]  \n",
       "2                                 [ajuin, atmoko]  \n",
       "3                            [authentic, algiers]  \n",
       "4                               [academy, aliano]  \n",
       "5                                    [abf, angke]  \n",
       "6                               [awapi, anterior]  \n",
       "7                            [acontoh, applikasi]  \n",
       "8                                [aqiel, arianto]  \n",
       "9                                 [affu, anandia]  \n",
       "10                            [aula, abdisatrijo]  \n",
       "11  [alamiahglucosaminechondroitinprotein, apart]  \n",
       "12                                 [abar, aliano]  \n",
       "13                          [aksan, alkilbenzena]  \n",
       "14                        [arofah, agentdipegang]  \n",
       "15                               [alive, apestak]  \n",
       "16                                 [abar, aliano]  \n",
       "17                              [affair, adailan]  \n",
       "18                       [arthaputra, abdurahman]  \n",
       "19                               [addendum, akar]  \n",
       "20                                 [aby, aslavia]  \n",
       "21              [arteriosklerosis, antarpersonel]  \n",
       "22                                   [asdep, aol]  \n",
       "23                     [aristoteles, aktifitasku]  \n",
       "24                              [amburadul, anis]  \n",
       "25                               [apkikasi, atak]  \n",
       "26                             [aremania, aureus]  \n",
       "27                          [associataed, armory]  \n",
       "28                                [ali, adhiguna]  "
      ]
     },
     "execution_count": 288,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.DataFrame(bigram_unigram)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3ca7661f",
   "metadata": {},
   "source": [
    "Because the case given is topic extraction, I think to better map out the algorithm to determine highly correlated words to a topic.\n",
    "If you look at the cleaning of the corpus that I did, it was still not optimal.\n",
    "It can be seen from the preprocessing that I did, some words still don't make sense to some topics, such as in the `Economy` class having the unigrams `Sarita` and `Lucy.`\n",
    "\n",
    "If given more time,\n",
    "I want to clean Indonesian text even better with the help of an external dataset to clean up existing slang (colloquial Indonesian lexicon) so that model can extract formal content and content with simple language created by the general public.\n",
    "\n",
    "Here I have not mapped the ROC and AUC metrics, so the model cannot be directly deployed because there can be many prediction errors if given new data. Besides that, I also want to try some more advanced machine learning models to compare with simple models"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "60c8c4ed",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python [conda env:python3.6]",
   "language": "python",
   "name": "conda-env-python3.6-py"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}