News_Title_Topic_Extraction.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "News Title Topic Extraction",
      "provenance": [],
      "collapsed_sections": [
        "A5v56Vkb5gB2"
      ],
      "toc_visible": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    }
  },
  "cells": [
    {
      "cell_type": "code",
      "metadata": {
        "id": "S_tqjLTVAVep",
        "colab_type": "code",
        "outputId": "b9d2e530-991f-4b8c-c322-798710a1cf30",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 197
        }
      },
      "source": [
        "try:\n",
        "    import Sastrawi\n",
        "except ModuleNotFoundError:\n",
        "    ! pip install PySastrawi\n",
        "finally:\n",
        "    from Sastrawi.StopWordRemover.StopWordRemoverFactory import StopWordRemoverFactory\n",
        "    from Sastrawi.Stemmer.StemmerFactory import StemmerFactory\n",
        "\n",
        "import os\n",
        "import pandas as pd\n",
        "\n",
        "if not os.path.exists('detik_news_title.csv'):\n",
        "    ! wget https://ibam.imfast.io/detik_news_title.csv\n",
        "\n",
        "\n",
        "df = pd.read_csv('detik_news_title.csv').drop(columns=['Unnamed: 0'])\n",
        "df.head()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>date</th>\n",
              "      <th>url</th>\n",
              "      <th>title</th>\n",
              "      <th>category</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>02/26/2020</td>\n",
              "      <td>https://finance.detik.com/berita-ekonomi-bisni...</td>\n",
              "      <td>Kemnaker Awasi TKA di Meikarta</td>\n",
              "      <td>finance</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>02/26/2020</td>\n",
              "      <td>https://finance.detik.com/berita-ekonomi-bisni...</td>\n",
              "      <td>BNI Digitalkan BNI Java Jazz 2020</td>\n",
              "      <td>finance</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>02/26/2020</td>\n",
              "      <td>https://finance.detik.com/berita-ekonomi-bisni...</td>\n",
              "      <td>Terbang ke Australia, Edhy Prabowo Mau Genjot ...</td>\n",
              "      <td>finance</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>02/26/2020</td>\n",
              "      <td>https://finance.detik.com/moneter/d-4916133/oj...</td>\n",
              "      <td>OJK Siapkan Stimulus Ekonomi Antisipasi Dampak...</td>\n",
              "      <td>finance</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>02/26/2020</td>\n",
              "      <td>https://finance.detik.com/berita-ekonomi-bisni...</td>\n",
              "      <td>Saran Buat Anies-RK yang Mangkir Rapat Banjir ...</td>\n",
              "      <td>finance</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "         date  ... category\n",
              "0  02/26/2020  ...  finance\n",
              "1  02/26/2020  ...  finance\n",
              "2  02/26/2020  ...  finance\n",
              "3  02/26/2020  ...  finance\n",
              "4  02/26/2020  ...  finance\n",
              "\n",
              "[5 rows x 4 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 68
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "tEO-V04Q8NX3",
        "colab_type": "text"
      },
      "source": [
        "## Word Count per Title"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "zVkCmdVN8Jsb",
        "colab_type": "code",
        "outputId": "b8f0523f-c080-471d-9f03-05fbdcf37f85",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 441
        }
      },
      "source": [
        "df['word_count'] = df['title'].apply(lambda x: len(str(x).split(' ')))\n",
        "df[['title', 'word_count']].head()\n",
        "df.hist(bins=20)\n",
        "df.word_count.describe()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "count    91017.000000\n",
              "mean         9.273674\n",
              "std          1.891406\n",
              "min          1.000000\n",
              "25%          8.000000\n",
              "50%          9.000000\n",
              "75%         11.000000\n",
              "max         20.000000\n",
              "Name: word_count, dtype: float64"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 69
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "2FXiELbQ9N93",
        "colab_type": "text"
      },
      "source": [
        "## Most Common and Uncommon Words per Category"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "wyKD_Go9-K64",
        "colab_type": "code",
        "outputId": "28a73a57-19e4-4fbd-b97e-0f5760b5b0dd",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 356
        }
      },
      "source": [
        "def get_top_words(df, n=10, column='title'):\n",
        "    top_words = pd.Series(' '.join(df[column]).split()).value_counts()[:n]\n",
        "    print(f'Top words {df.category.unique()[0]} category: ', end='')\n",
        "    for index, value in top_words.items():\n",
        "        print(f'{index} ({value}), ', end='')\n",
        "    print('\\n')\n",
        "\n",
        "for cat in df.category.unique():\n",
        "    get_top_words(df[df.category == cat])"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Top words finance category: di (2555), Corona (1645), Ini (1465), Rp (1387), ke (1116), RI (951), yang (805), dan (753), Bisa (703), Corona, (699), \n",
            "\n",
            "Top words food category: Ini (1690), di (1253), yang (1091), 5 (847), dan (814), Makanan (673), Makan (521), Foto (458), Galeri (438), hingga (415), \n",
            "\n",
            "Top words health category: Corona (2064), di (1275), Virus (1026), Ini (1024), yang (713), Pasien (472), Bisa (431), Corona, (385), dan (372), untuk (369), \n",
            "\n",
            "Top words hot category: di (3100), dan (1649), yang (1365), Ini (1197), Tak (1076), Jadi (1001), Corona (997), dengan (913), dari (799), ke (672), \n",
            "\n",
            "Top words inet category: di (1076), Corona (665), Ini (543), dan (518), yang (494), Virus (336), Ponsel (298), dari (267), untuk (258), Bikin (253), \n",
            "\n",
            "Top words news category: di (13632), Corona (5488), yang (2495), ke (2342), Tak (2235), Ini (2101), Kasus (1938), Warga (1881), Polisi (1854), dan (1782), \n",
            "\n",
            "Top words oto category: Mobil (1125), di (957), Ini (669), yang (483), Motor (414), dan (318), Corona (316), Rp (267), Honda (239), Corona, (226), \n",
            "\n",
            "Top words sport category: di (508), 2020 (287), Indonesia (213), MotoGP (194), ke (179), dan (162), Ini (142), Corona (121), Virus (118), Tak (118), \n",
            "\n",
            "Top words travel category: di (2499), yang (1001), Ini (907), Corona (687), Wisata (509), ke (434), dan (418), dari (405), Corona, (313), Turis (274), \n",
            "\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "g2CiqEkSADVX",
        "colab_type": "text"
      },
      "source": [
        "## Text Preprocessing"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "w13ZR7MGAHbI",
        "colab_type": "code",
        "outputId": "92908859-24a1-4624-d180-df03f8521716",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        }
      },
      "source": [
        "import re\n",
        "def preprocess(text):\n",
        "    # remove punctuation\n",
        "    text = re.sub('[^a-zA-Z]', ' ', text)\n",
        "    # convert to lower case\n",
        "    text = text.lower()\n",
        "    # remove special char and digit\n",
        "    text=re.sub(\"(\\\\d|\\\\W)+\",\" \",text)\n",
        "    # remove stop words\n",
        "    stopword = StopWordRemoverFactory().create_stop_word_remover() \n",
        "    text = stopword.remove(text)\n",
        "    # stemming\n",
        "    stemmer = StemmerFactory().create_stemmer()\n",
        "    text = stemmer.stem(text)\n",
        "    return text\n",
        "\n",
        "import time\n",
        "\n",
        "start_time = time.time()\n",
        "for i in range(0, len(df)):\n",
        "    if (i%500) == 0:\n",
        "        mid_time = time.time()\n",
        "        print(f'Processing #{i} data, {(mid_time - start_time):.2f} seconds elapsed.')\n",
        "\n",
        "    df.at[i, 'corpus'] = preprocess(df['title'][i])\n",
        "\n",
        "end_time = time.time()\n",
        "print(f'End processing {len(df)} data for {(end_time-start_time):.2f} seconds.')\n",
        "df.head(15)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Processing #0 data, 0.00 seconds elapsed.\n",
            "Processing #500 data, 3.46 seconds elapsed.\n",
            "Processing #1000 data, 6.91 seconds elapsed.\n",
            "Processing #1500 data, 10.27 seconds elapsed.\n",
            "Processing #2000 data, 13.88 seconds elapsed.\n",
            "Processing #2500 data, 17.11 seconds elapsed.\n",
            "Processing #3000 data, 20.61 seconds elapsed.\n",
            "Processing #3500 data, 24.00 seconds elapsed.\n",
            "Processing #4000 data, 27.43 seconds elapsed.\n",
            "Processing #4500 data, 30.83 seconds elapsed.\n",
            "Processing #5000 data, 34.20 seconds elapsed.\n",
            "Processing #5500 data, 37.62 seconds elapsed.\n",
            "Processing #6000 data, 40.89 seconds elapsed.\n",
            "Processing #6500 data, 44.27 seconds elapsed.\n",
            "Processing #7000 data, 47.64 seconds elapsed.\n",
            "Processing #7500 data, 50.99 seconds elapsed.\n",
            "Processing #8000 data, 54.61 seconds elapsed.\n",
            "Processing #8500 data, 58.21 seconds elapsed.\n",
            "Processing #9000 data, 61.71 seconds elapsed.\n",
            "Processing #9500 data, 65.18 seconds elapsed.\n",
            "Processing #10000 data, 68.73 seconds elapsed.\n",
            "Processing #10500 data, 72.19 seconds elapsed.\n",
            "Processing #11000 data, 75.67 seconds elapsed.\n",
            "Processing #11500 data, 79.18 seconds elapsed.\n",
            "Processing #12000 data, 82.65 seconds elapsed.\n",
            "Processing #12500 data, 86.12 seconds elapsed.\n",
            "Processing #13000 data, 89.73 seconds elapsed.\n",
            "Processing #13500 data, 93.15 seconds elapsed.\n",
            "Processing #14000 data, 96.63 seconds elapsed.\n",
            "Processing #14500 data, 100.05 seconds elapsed.\n",
            "Processing #15000 data, 103.59 seconds elapsed.\n",
            "Processing #15500 data, 107.05 seconds elapsed.\n",
            "Processing #16000 data, 110.79 seconds elapsed.\n",
            "Processing #16500 data, 114.34 seconds elapsed.\n",
            "Processing #17000 data, 117.78 seconds elapsed.\n",
            "Processing #17500 data, 121.30 seconds elapsed.\n",
            "Processing #18000 data, 124.71 seconds elapsed.\n",
            "Processing #18500 data, 128.14 seconds elapsed.\n",
            "Processing #19000 data, 131.57 seconds elapsed.\n",
            "Processing #19500 data, 135.04 seconds elapsed.\n",
            "Processing #20000 data, 138.53 seconds elapsed.\n",
            "Processing #20500 data, 142.08 seconds elapsed.\n",
            "Processing #21000 data, 145.57 seconds elapsed.\n",
            "Processing #21500 data, 149.06 seconds elapsed.\n",
            "Processing #22000 data, 152.60 seconds elapsed.\n",
            "Processing #22500 data, 156.12 seconds elapsed.\n",
            "Processing #23000 data, 159.87 seconds elapsed.\n",
            "Processing #23500 data, 163.73 seconds elapsed.\n",
            "Processing #24000 data, 167.71 seconds elapsed.\n",
            "Processing #24500 data, 171.69 seconds elapsed.\n",
            "Processing #25000 data, 175.40 seconds elapsed.\n",
            "Processing #25500 data, 179.04 seconds elapsed.\n",
            "Processing #26000 data, 182.95 seconds elapsed.\n",
            "Processing #26500 data, 186.94 seconds elapsed.\n",
            "Processing #27000 data, 190.91 seconds elapsed.\n",
            "Processing #27500 data, 194.77 seconds elapsed.\n",
            "Processing #28000 data, 198.50 seconds elapsed.\n",
            "Processing #28500 data, 202.17 seconds elapsed.\n",
            "Processing #29000 data, 205.86 seconds elapsed.\n",
            "Processing #29500 data, 209.57 seconds elapsed.\n",
            "Processing #30000 data, 213.32 seconds elapsed.\n",
            "Processing #30500 data, 217.03 seconds elapsed.\n",
            "Processing #31000 data, 220.78 seconds elapsed.\n",
            "Processing #31500 data, 224.51 seconds elapsed.\n",
            "Processing #32000 data, 228.28 seconds elapsed.\n",
            "Processing #32500 data, 231.94 seconds elapsed.\n",
            "Processing #33000 data, 235.79 seconds elapsed.\n",
            "Processing #33500 data, 239.50 seconds elapsed.\n",
            "Processing #34000 data, 243.31 seconds elapsed.\n",
            "Processing #34500 data, 247.01 seconds elapsed.\n",
            "Processing #35000 data, 250.71 seconds elapsed.\n",
            "Processing #35500 data, 254.39 seconds elapsed.\n",
            "Processing #36000 data, 258.14 seconds elapsed.\n",
            "Processing #36500 data, 261.87 seconds elapsed.\n",
            "Processing #37000 data, 265.64 seconds elapsed.\n",
            "Processing #37500 data, 269.33 seconds elapsed.\n",
            "Processing #38000 data, 273.04 seconds elapsed.\n",
            "Processing #38500 data, 276.76 seconds elapsed.\n",
            "Processing #39000 data, 280.55 seconds elapsed.\n",
            "Processing #39500 data, 284.60 seconds elapsed.\n",
            "Processing #40000 data, 288.29 seconds elapsed.\n",
            "Processing #40500 data, 291.87 seconds elapsed.\n",
            "Processing #41000 data, 295.49 seconds elapsed.\n",
            "Processing #41500 data, 299.23 seconds elapsed.\n",
            "Processing #42000 data, 302.93 seconds elapsed.\n",
            "Processing #42500 data, 306.83 seconds elapsed.\n",
            "Processing #43000 data, 310.64 seconds elapsed.\n",
            "Processing #43500 data, 314.42 seconds elapsed.\n",
            "Processing #44000 data, 318.08 seconds elapsed.\n",
            "Processing #44500 data, 321.72 seconds elapsed.\n",
            "Processing #45000 data, 325.30 seconds elapsed.\n",
            "Processing #45500 data, 328.95 seconds elapsed.\n",
            "Processing #46000 data, 332.65 seconds elapsed.\n",
            "Processing #46500 data, 336.37 seconds elapsed.\n",
            "Processing #47000 data, 340.52 seconds elapsed.\n",
            "Processing #47500 data, 344.38 seconds elapsed.\n",
            "Processing #48000 data, 348.07 seconds elapsed.\n",
            "Processing #48500 data, 351.77 seconds elapsed.\n",
            "Processing #49000 data, 355.52 seconds elapsed.\n",
            "Processing #49500 data, 359.35 seconds elapsed.\n",
            "Processing #50000 data, 363.13 seconds elapsed.\n",
            "Processing #50500 data, 367.13 seconds elapsed.\n",
            "Processing #51000 data, 371.25 seconds elapsed.\n",
            "Processing #51500 data, 375.24 seconds elapsed.\n",
            "Processing #52000 data, 379.05 seconds elapsed.\n",
            "Processing #52500 data, 382.79 seconds elapsed.\n",
            "Processing #53000 data, 386.66 seconds elapsed.\n",
            "Processing #53500 data, 390.29 seconds elapsed.\n",
            "Processing #54000 data, 393.89 seconds elapsed.\n",
            "Processing #54500 data, 397.69 seconds elapsed.\n",
            "Processing #55000 data, 401.38 seconds elapsed.\n",
            "Processing #55500 data, 405.21 seconds elapsed.\n",
            "Processing #56000 data, 409.06 seconds elapsed.\n",
            "Processing #56500 data, 412.85 seconds elapsed.\n",
            "Processing #57000 data, 416.73 seconds elapsed.\n",
            "Processing #57500 data, 420.54 seconds elapsed.\n",
            "Processing #58000 data, 424.36 seconds elapsed.\n",
            "Processing #58500 data, 428.12 seconds elapsed.\n",
            "Processing #59000 data, 431.87 seconds elapsed.\n",
            "Processing #59500 data, 435.65 seconds elapsed.\n",
            "Processing #60000 data, 439.61 seconds elapsed.\n",
            "Processing #60500 data, 443.69 seconds elapsed.\n",
            "Processing #61000 data, 447.36 seconds elapsed.\n",
            "Processing #61500 data, 451.24 seconds elapsed.\n",
            "Processing #62000 data, 454.89 seconds elapsed.\n",
            "Processing #62500 data, 458.61 seconds elapsed.\n",
            "Processing #63000 data, 462.16 seconds elapsed.\n",
            "Processing #63500 data, 465.77 seconds elapsed.\n",
            "Processing #64000 data, 469.25 seconds elapsed.\n",
            "Processing #64500 data, 472.68 seconds elapsed.\n",
            "Processing #65000 data, 476.22 seconds elapsed.\n",
            "Processing #65500 data, 479.69 seconds elapsed.\n",
            "Processing #66000 data, 483.14 seconds elapsed.\n",
            "Processing #66500 data, 486.66 seconds elapsed.\n",
            "Processing #67000 data, 490.19 seconds elapsed.\n",
            "Processing #67500 data, 493.75 seconds elapsed.\n",
            "Processing #68000 data, 497.63 seconds elapsed.\n",
            "Processing #68500 data, 501.47 seconds elapsed.\n",
            "Processing #69000 data, 505.40 seconds elapsed.\n",
            "Processing #69500 data, 509.14 seconds elapsed.\n",
            "Processing #70000 data, 512.99 seconds elapsed.\n",
            "Processing #70500 data, 517.06 seconds elapsed.\n",
            "Processing #71000 data, 520.85 seconds elapsed.\n",
            "Processing #71500 data, 524.75 seconds elapsed.\n",
            "Processing #72000 data, 528.40 seconds elapsed.\n",
            "Processing #72500 data, 532.18 seconds elapsed.\n",
            "Processing #73000 data, 536.01 seconds elapsed.\n",
            "Processing #73500 data, 539.79 seconds elapsed.\n",
            "Processing #74000 data, 543.53 seconds elapsed.\n",
            "Processing #74500 data, 547.23 seconds elapsed.\n",
            "Processing #75000 data, 550.94 seconds elapsed.\n",
            "Processing #75500 data, 554.64 seconds elapsed.\n",
            "Processing #76000 data, 558.62 seconds elapsed.\n",
            "Processing #76500 data, 562.42 seconds elapsed.\n",
            "Processing #77000 data, 566.16 seconds elapsed.\n",
            "Processing #77500 data, 570.03 seconds elapsed.\n",
            "Processing #78000 data, 573.63 seconds elapsed.\n",
            "Processing #78500 data, 577.27 seconds elapsed.\n",
            "Processing #79000 data, 580.93 seconds elapsed.\n",
            "Processing #79500 data, 584.68 seconds elapsed.\n",
            "Processing #80000 data, 588.26 seconds elapsed.\n",
            "Processing #80500 data, 591.82 seconds elapsed.\n",
            "Processing #81000 data, 595.43 seconds elapsed.\n",
            "Processing #81500 data, 599.06 seconds elapsed.\n",
            "Processing #82000 data, 602.70 seconds elapsed.\n",
            "Processing #82500 data, 606.41 seconds elapsed.\n",
            "Processing #83000 data, 610.06 seconds elapsed.\n",
            "Processing #83500 data, 614.10 seconds elapsed.\n",
            "Processing #84000 data, 618.28 seconds elapsed.\n",
            "Processing #84500 data, 622.25 seconds elapsed.\n",
            "Processing #85000 data, 626.04 seconds elapsed.\n",
            "Processing #85500 data, 629.53 seconds elapsed.\n",
            "Processing #86000 data, 633.02 seconds elapsed.\n",
            "Processing #86500 data, 636.54 seconds elapsed.\n",
            "Processing #87000 data, 640.04 seconds elapsed.\n",
            "Processing #87500 data, 643.66 seconds elapsed.\n",
            "Processing #88000 data, 647.44 seconds elapsed.\n",
            "Processing #88500 data, 650.96 seconds elapsed.\n",
            "Processing #89000 data, 654.52 seconds elapsed.\n",
            "Processing #89500 data, 658.14 seconds elapsed.\n",
            "Processing #90000 data, 661.92 seconds elapsed.\n",
            "Processing #90500 data, 665.55 seconds elapsed.\n",
            "Processing #91000 data, 669.53 seconds elapsed.\n",
            "End processing 91017 data for 669.66 seconds.\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>date</th>\n",
              "      <th>url</th>\n",
              "      <th>title</th>\n",
              "      <th>category</th>\n",
              "      <th>word_count</th>\n",
              "      <th>corpus</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>02/26/2020</td>\n",
              "      <td>https://finance.detik.com/berita-ekonomi-bisni...</td>\n",
              "      <td>Kemnaker Awasi TKA di Meikarta</td>\n",
              "      <td>finance</td>\n",
              "      <td>5</td>\n",
              "      <td>kemnaker awas tka meikarta</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>02/26/2020</td>\n",
              "      <td>https://finance.detik.com/berita-ekonomi-bisni...</td>\n",
              "      <td>BNI Digitalkan BNI Java Jazz 2020</td>\n",
              "      <td>finance</td>\n",
              "      <td>6</td>\n",
              "      <td>bni digital bni java jazz</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>02/26/2020</td>\n",
              "      <td>https://finance.detik.com/berita-ekonomi-bisni...</td>\n",
              "      <td>Terbang ke Australia, Edhy Prabowo Mau Genjot ...</td>\n",
              "      <td>finance</td>\n",
              "      <td>10</td>\n",
              "      <td>terbang australia edhy prabowo genjot budi day...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>02/26/2020</td>\n",
              "      <td>https://finance.detik.com/moneter/d-4916133/oj...</td>\n",
              "      <td>OJK Siapkan Stimulus Ekonomi Antisipasi Dampak...</td>\n",
              "      <td>finance</td>\n",
              "      <td>7</td>\n",
              "      <td>ojk siap stimulus ekonomi antisipasi dampak co...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>02/26/2020</td>\n",
              "      <td>https://finance.detik.com/berita-ekonomi-bisni...</td>\n",
              "      <td>Saran Buat Anies-RK yang Mangkir Rapat Banjir ...</td>\n",
              "      <td>finance</td>\n",
              "      <td>9</td>\n",
              "      <td>saran anies rk mangkir rapat banjir dpr</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>02/26/2020</td>\n",
              "      <td>https://finance.detik.com/berita-ekonomi-bisni...</td>\n",
              "      <td>RI Bakal Punya Destinasi Wisata Sejarah Islam ...</td>\n",
              "      <td>finance</td>\n",
              "      <td>9</td>\n",
              "      <td>ri destinasi wisata sejarah islam ancol</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6</th>\n",
              "      <td>02/26/2020</td>\n",
              "      <td>https://finance.detik.com/perencanaan-keuangan...</td>\n",
              "      <td>Cara Biar Nggak Tertipu Travel Umrah Abal-abal...</td>\n",
              "      <td>finance</td>\n",
              "      <td>10</td>\n",
              "      <td>biar nggak tipu travel umrah abal abal first t...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7</th>\n",
              "      <td>02/26/2020</td>\n",
              "      <td>https://finance.detik.com/berita-ekonomi-bisni...</td>\n",
              "      <td>DPR Kritik Cara Anies Atasi Banjir, Pesan Kera...</td>\n",
              "      <td>finance</td>\n",
              "      <td>12</td>\n",
              "      <td>dpr kritik anies atas banjir pesan keras erick...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>8</th>\n",
              "      <td>02/26/2020</td>\n",
              "      <td>https://finance.detik.com/moneter/d-4916086/ja...</td>\n",
              "      <td>Jadi Korban Gagal Bayar Asuransi? Bisa Mengadu...</td>\n",
              "      <td>finance</td>\n",
              "      <td>9</td>\n",
              "      <td>korban gagal bayar asuransi adu</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9</th>\n",
              "      <td>02/26/2020</td>\n",
              "      <td>https://finance.detik.com/berita-ekonomi-bisni...</td>\n",
              "      <td>Erick Thohir: BUMN Bukan Badan Usaha Milik Nen...</td>\n",
              "      <td>finance</td>\n",
              "      <td>9</td>\n",
              "      <td>erick thohir bumn badan usaha milik nenek lu</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>10</th>\n",
              "      <td>02/26/2020</td>\n",
              "      <td>https://finance.detik.com/berita-ekonomi-bisni...</td>\n",
              "      <td>Terbitkan Resi Pertama, SRG Konawe Bantu Ketah...</td>\n",
              "      <td>finance</td>\n",
              "      <td>9</td>\n",
              "      <td>terbit resi srg konawe bantu tahan pangan nasi...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>11</th>\n",
              "      <td>02/26/2020</td>\n",
              "      <td>https://finance.detik.com/properti/d-4915865/j...</td>\n",
              "      <td>Jokowi Janji Buka 3 Juta Lapangan Kerja di Ibu...</td>\n",
              "      <td>finance</td>\n",
              "      <td>12</td>\n",
              "      <td>jokowi janji buka juta lapang kota negara</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>12</th>\n",
              "      <td>02/26/2020</td>\n",
              "      <td>https://finance.detik.com/detiktv/d-4916004/vi...</td>\n",
              "      <td>Video Erick Thohir: Banyak Pihak Swasta Akali ...</td>\n",
              "      <td>finance</td>\n",
              "      <td>8</td>\n",
              "      <td>video erick thohir swasta akal bumn</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>13</th>\n",
              "      <td>02/26/2020</td>\n",
              "      <td>https://finance.detik.com/berita-ekonomi-bisni...</td>\n",
              "      <td>Wishnutama Mau Sewa Influencer Asing, Pakai BTS?</td>\n",
              "      <td>finance</td>\n",
              "      <td>7</td>\n",
              "      <td>wishnutama sewa influencer asing pakai bts</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>14</th>\n",
              "      <td>02/26/2020</td>\n",
              "      <td>https://finance.detik.com/berita-ekonomi-bisni...</td>\n",
              "      <td>Dikritik Keras DPR soal Banjir, Anies Buka Suara</td>\n",
              "      <td>finance</td>\n",
              "      <td>8</td>\n",
              "      <td>kritik keras dpr banjir anies buka suara</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "          date  ...                                             corpus\n",
              "0   02/26/2020  ...                         kemnaker awas tka meikarta\n",
              "1   02/26/2020  ...                          bni digital bni java jazz\n",
              "2   02/26/2020  ...  terbang australia edhy prabowo genjot budi day...\n",
              "3   02/26/2020  ...  ojk siap stimulus ekonomi antisipasi dampak co...\n",
              "4   02/26/2020  ...            saran anies rk mangkir rapat banjir dpr\n",
              "5   02/26/2020  ...            ri destinasi wisata sejarah islam ancol\n",
              "6   02/26/2020  ...  biar nggak tipu travel umrah abal abal first t...\n",
              "7   02/26/2020  ...  dpr kritik anies atas banjir pesan keras erick...\n",
              "8   02/26/2020  ...                    korban gagal bayar asuransi adu\n",
              "9   02/26/2020  ...       erick thohir bumn badan usaha milik nenek lu\n",
              "10  02/26/2020  ...  terbit resi srg konawe bantu tahan pangan nasi...\n",
              "11  02/26/2020  ...          jokowi janji buka juta lapang kota negara\n",
              "12  02/26/2020  ...                video erick thohir swasta akal bumn\n",
              "13  02/26/2020  ...         wishnutama sewa influencer asing pakai bts\n",
              "14  02/26/2020  ...           kritik keras dpr banjir anies buka suara\n",
              "\n",
              "[15 rows x 6 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 91
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "0ihugXI95SHv",
        "colab_type": "text"
      },
      "source": [
        "## Extracting the Topic using TF-IDF"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Qt0Q4S6q5ZzE",
        "colab_type": "text"
      },
      "source": [
        "### Create a word count vector using CountVectorizer"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Ob8T6RU05fr6",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from sklearn.feature_extraction.text import CountVectorizer\n",
        "cv=CountVectorizer(max_df=0.8, max_features=10000, ngram_range=(1,2))\n",
        "X=cv.fit_transform(df['corpus'])"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "A5v56Vkb5gB2",
        "colab_type": "text"
      },
      "source": [
        "#### Visualize Top Word (Single n-gram and bigram)"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Y7KKEp7mMY8x",
        "colab_type": "code",
        "outputId": "fb9494f8-f999-4b31-b24e-18aa5a488158",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 502
        }
      },
      "source": [
        "from sklearn.feature_extraction.text import CountVectorizer\n",
        "\n",
        "#Most frequently occuring words\n",
        "def get_top_n_words(corpus, n=None):\n",
        "    vec = CountVectorizer().fit(corpus)\n",
        "    bag_of_words = vec.transform(corpus)\n",
        "    sum_words = bag_of_words.sum(axis=0) \n",
        "    words_freq = [(word, sum_words[0, idx]) for word, idx in      \n",
        "                   vec.vocabulary_.items()]\n",
        "    words_freq =sorted(words_freq, key = lambda x: x[1], \n",
        "                       reverse=True)\n",
        "    return words_freq[:n]\n",
        "#Convert most freq words to dataframe for plotting bar plot\n",
        "top_words = get_top_n_words(df['corpus'], n=10)\n",
        "top_df = pd.DataFrame(top_words)\n",
        "top_df.columns=[\"Word\", \"Freq\"]\n",
        "# #Barplot of most freq words\n",
        "import seaborn as sns\n",
        "sns.set(rc={'figure.figsize':(15,8)})\n",
        "g = sns.barplot(x=\"Freq\", y=\"Word\", data=top_df)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1080x576 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "I5a0ZeywNQKY",
        "colab_type": "code",
        "outputId": "ecd40208-f689-4753-d69f-2b33e59a6ba6",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 502
        }
      },
      "source": [
        "#Most frequently occuring Bi-grams\n",
        "def get_top_n2_words(corpus, n=None):\n",
        "    vec1 = CountVectorizer(ngram_range=(5,5),  \n",
        "            max_features=2000).fit(corpus)\n",
        "    bag_of_words = vec1.transform(corpus)\n",
        "    sum_words = bag_of_words.sum(axis=0) \n",
        "    words_freq = [(word, sum_words[0, idx]) for word, idx in     \n",
        "                  vec1.vocabulary_.items()]\n",
        "    words_freq =sorted(words_freq, key = lambda x: x[1], \n",
        "                reverse=True)\n",
        "    return words_freq[:n]\n",
        "top2_words = get_top_n2_words(df['corpus'], n=10)\n",
        "top2_df = pd.DataFrame(top2_words)\n",
        "top2_df.columns=[\"Bi-gram\", \"Freq\"]\n",
        "#Barplot of most freq Bi-grams\n",
        "import seaborn as sns\n",
        "sns.set(rc={'figure.figsize':(13,8)})\n",
        "h=sns.barplot(x=\"Freq\", y=\"Bi-gram\", data=top2_df)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 936x576 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "o8LYWg8j5pTL",
        "colab_type": "text"
      },
      "source": [
        "## Create TF-IDF Vector using vector of wordcount"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "yNR94TmjgVyQ",
        "colab_type": "code",
        "outputId": "72f23f1e-6b31-4f65-a3f9-e75a6e06c3f4",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "from sklearn.feature_extraction.text import TfidfTransformer\n",
        "tfidf_transformer=TfidfTransformer(smooth_idf=True,use_idf=True)\n",
        "tfidf_transformer.fit(X)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "TfidfTransformer(norm='l2', smooth_idf=True, sublinear_tf=False, use_idf=True)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 179
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "tanCo9woh9bu",
        "colab_type": "text"
      },
      "source": [
        "## Extracting Topic from Random Title"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "WiyAWxem6ZSz",
        "colab_type": "code",
        "outputId": "761c6caf-fb14-4313-fe2d-ff1a9169d7c7",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 194
        }
      },
      "source": [
        "feature_names=cv.get_feature_names()\n",
        "b\n",
        "#Function for sorting tf_idf in descending order\n",
        "from scipy.sparse import coo_matrix\n",
        "def sort_coo(coo_matrix):\n",
        "    tuples = zip(coo_matrix.col, coo_matrix.data)\n",
        "    return sorted(tuples, key=lambda x: (x[1], x[0]), reverse=True)\n",
        " \n",
        "def extract_topn_from_vector(feature_names, sorted_items, topn=10):\n",
        "    \"\"\"get the feature names and tf-idf score of top n items\"\"\"\n",
        "    \n",
        "    #use only topn items from vector\n",
        "    sorted_items = sorted_items[:topn]\n",
        " \n",
        "    score_vals = []\n",
        "    feature_vals = []\n",
        "    \n",
        "    # word index and corresponding tf-idf score\n",
        "    for idx, score in sorted_items:\n",
        "        \n",
        "        #keep track of feature name and its corresponding score\n",
        "        score_vals.append(round(score, 3))\n",
        "        feature_vals.append(feature_names[idx])\n",
        " \n",
        "    #create a tuples of feature,score\n",
        "    #results = zip(feature_vals,score_vals)\n",
        "    results= {}\n",
        "    for idx in range(len(feature_vals)):\n",
        "        results[feature_vals[idx]]=score_vals[idx]\n",
        "    \n",
        "    return results\n",
        "#sort the tf-idf vectors by descending order of scores\n",
        "sorted_items=sort_coo(tf_idf_vector.tocoo())\n",
        "#extract only the top n; n here is 10\n",
        "keywords=extract_topn_from_vector(feature_names,sorted_items,5)\n",
        " \n",
        "# now print the results\n",
        "print(\"\\nAbstract:\")\n",
        "print(doc)\n",
        "print(\"\\nKeywords:\")\n",
        "for k in keywords:\n",
        "    print(k,keywords[k])"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "\n",
            "Abstract:\n",
            "Pengakuan Kepsek di Bandung Kerap Dihubungi 'Titip Siswa' Saat PPDB\n",
            "\n",
            "Keywords:\n",
            "ppdb 0.459\n",
            "kepsek 0.44\n",
            "titip 0.438\n",
            "kerap 0.397\n",
            "siswa 0.356\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "8O92hzQIiEig",
        "colab_type": "text"
      },
      "source": [
        "Credits:\n",
        "https://medium.com/analytics-vidhya/automated-keyword-extraction-from-articles-using-nlp-bfd864f41b34"
      ]
    }
  ]
}