exploration.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "exploration.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "toc_visible": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    }
  },
  "cells": [
    {
      "cell_type": "code",
      "metadata": {
        "id": "x789RVTUKD3T",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "!pip install -q pythainlp\n",
        "import pandas as pd\n",
        "from sklearn.feature_extraction.text import CountVectorizer\n",
        "from pythainlp import word_tokenize\n",
        "from pythainlp import sent_tokenize\n",
        "import os\n",
        "from tqdm.notebook import tqdm\n",
        "import matplotlib\n",
        "import matplotlib.pyplot as plt\n",
        "import numpy as np\n",
        "plt.rcParams[\"font.family\"] = \"serif\""
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "zVkbl9ScKM6i",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "df_trtpbs = pd.read_csv('/content/drive/My Drive/Final_Datasets/tr-tpbs.csv')\n",
        "df_tr = pd.read_csv('/content/drive/My Drive/Final_Datasets/additional-datasets/thairath-222.csv')\n",
        "df_tpbs = pd.read_csv('/content/drive/My Drive/Final_Datasets/additional-datasets/thaiPBS-111k.csv')\n",
        "df_th2en = pd.read_csv('/content/drive/My Drive/Final_Datasets/TH-CLS-Corpus/trainset_th2en.csv')"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "MDBL2NJKKbdw",
        "colab_type": "text"
      },
      "source": [
        "# The Size Comparision of Datasets"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "vGJCDbD8Kd1t",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 331
        },
        "outputId": "fd85fb68-408d-4b32-a8cc-e2e28d475b22"
      },
      "source": [
        "objects = ('ThaiPBS-111K','Thairath-222K', 'TR-TPBS')\n",
        "y_pos = np.arange(len(objects))\n",
        "datasets_size = [len(df_tpbs), len(df_tr), len(df_trtpbs)]\n",
        "plt.bar(y_pos, datasets_size, align='center', alpha=1, width=0.4)\n",
        "plt.xticks(y_pos, objects)\n",
        "plt.ylabel('No. of Articles')\n",
        "plt.title('The Size Comparision of Datasets')\n",
        "plt.show()\n",
        "\n",
        "print(\"ThaiPBS-111k  contains \", len(df_tpbs), \" news\")\n",
        "print(\"Thairath-222k contains \", len(df_tr), \" news\")\n",
        "print(\"TR-TPBS       contains \", len(df_trtpbs), \" news\")"
      ],
      "execution_count": 95,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "text": [
            "ThaiPBS-111k  contains  111614  news\n",
            "Thairath-222k contains  222553  news\n",
            "TR-TPBS       contains  310926  news\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "bNoDBCwkNcQ8",
        "colab_type": "text"
      },
      "source": [
        "#Elements (Columns) Contained in each Dataset"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "RJjSYtH_NQAH",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 165
        },
        "outputId": "fe47d2c6-c000-4f9e-9d34-1feab3d99ff4"
      },
      "source": [
        "print('TR-TPBS')\n",
        "print('Number of Aritcles: ', len(df_trtpbs))\n",
        "df_trtpbs.head(1)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "TR-TPBS\n",
            "Number of Aritcles:  310926\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>body</th>\n",
              "      <th>labels</th>\n",
              "      <th>summary</th>\n",
              "      <th>tags</th>\n",
              "      <th>title</th>\n",
              "      <th>url</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>วันนี้ (23 ม.ค. 2559) นายธวัชชัย จิตรภาษ์นันท์...</td>\n",
              "      <td>เศรษฐกิจ</td>\n",
              "      <td>กสทช. เตรียมขอการสนับสนุนงบประมาณจากกองทุน กทป...</td>\n",
              "      <td>ระบบจัดเรตติ้งโทรทัศน์ไทยใหม่,กสทช.,กองทุน กทป...</td>\n",
              "      <td>กสทช.เตรียมของบ กทปส. ปฏิรูปวิจัยเรตติ้งทีวีแบ...</td>\n",
              "      <td>https://news.thaipbs.or.th/content/7496</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "                                                body  ...                                      url\n",
              "0  วันนี้ (23 ม.ค. 2559) นายธวัชชัย จิตรภาษ์นันท์...  ...  https://news.thaipbs.or.th/content/7496\n",
              "\n",
              "[1 rows x 6 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 8
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "vGL-kLPLNf17",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 270
        },
        "outputId": "20c1ef01-7435-4e28-ad42-b8ea5956fc8a"
      },
      "source": [
        "print('Thairath-222K')\n",
        "print('Number of Aritcles: ', len(df_tr))\n",
        "df_tr.head(1) "
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Thairath-222K\n",
            "Number of Aritcles:  222553\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>title</th>\n",
              "      <th>body</th>\n",
              "      <th>summary</th>\n",
              "      <th>type</th>\n",
              "      <th>tags</th>\n",
              "      <th>url</th>\n",
              "      <th>date</th>\n",
              "      <th>title_length</th>\n",
              "      <th>article_length</th>\n",
              "      <th>summary_length</th>\n",
              "      <th>abs_n1</th>\n",
              "      <th>abs_n2</th>\n",
              "      <th>abs_n3</th>\n",
              "      <th>abs_n4</th>\n",
              "      <th>abs_n5</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>โป้งเดียวร่วง \"ฝนพันล้าน\" ต่อยแจ๊คแพ้น็อก 4 มว...</td>\n",
              "      <td>วันที่ 30 พ.ย. 62  ศึกมวยไทยเกียรติเพชร ซูเปอร...</td>\n",
              "      <td>ฝนพันล้าน พี.เค.แสนชัยมวยไทยยิม หมัดหนักเอาเรื...</td>\n",
              "      <td>กีฬา,มวย/MMA</td>\n",
              "      <td>มวยลุมพินี,มวยไทย,ฝนพันล้าน พี.เค.แสนชัยมวยไทย...</td>\n",
              "      <td>https://www.thairath.co.th/sport/boxing-mma/17...</td>\n",
              "      <td>2019-12-01T00:30:00+07:00</td>\n",
              "      <td>16.0</td>\n",
              "      <td>248.0</td>\n",
              "      <td>43.0</td>\n",
              "      <td>20.0</td>\n",
              "      <td>33.333333</td>\n",
              "      <td>41.463415</td>\n",
              "      <td>50.0</td>\n",
              "      <td>58.974359</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "                                               title  ...     abs_n5\n",
              "0  โป้งเดียวร่วง \"ฝนพันล้าน\" ต่อยแจ๊คแพ้น็อก 4 มว...  ...  58.974359\n",
              "\n",
              "[1 rows x 15 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 9
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Rj4Zi4RwNiZr",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 287
        },
        "outputId": "a0f8ba8c-5e16-45d8-cb3d-4e0ab8500efb"
      },
      "source": [
        "print('ThaiPBS-111K')\n",
        "print('Number of Aritcles: ', len(df_tpbs))\n",
        "df_tpbs.head(1)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "ThaiPBS-111K\n",
            "Number of Aritcles:  111614\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>title</th>\n",
              "      <th>body</th>\n",
              "      <th>summary</th>\n",
              "      <th>tags</th>\n",
              "      <th>label</th>\n",
              "      <th>url</th>\n",
              "      <th>article_length</th>\n",
              "      <th>summary_length</th>\n",
              "      <th>title_length</th>\n",
              "      <th>abstractedness_n1</th>\n",
              "      <th>abstractedness_n2</th>\n",
              "      <th>abstractedness_n3</th>\n",
              "      <th>abstractedness_n4</th>\n",
              "      <th>abstractedness_n5</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>นักวิจัยออสเตรเลียเผยสาเหตุฉลามโจมตีมนุษย์</td>\n",
              "      <td>นายนาธาน ฮาร์ท หัวหน้าทีมวิจัยจากมหาวิทยาลัยเว...</td>\n",
              "      <td>ทีมนักวิจัยออสเตรเลียศึกษาว่าเหตุใดฉลามถึงจู่...</td>\n",
              "      <td>ฉลาม,นักวิจัย,มนุษย์,ออสเตรเลีย,แมวน้ำ,โจมตี</td>\n",
              "      <td>ต่างประเทศ</td>\n",
              "      <td>https://news.thaipbs.or.th/content/1</td>\n",
              "      <td>148.0</td>\n",
              "      <td>23.0</td>\n",
              "      <td>7.0</td>\n",
              "      <td>4.761905</td>\n",
              "      <td>50.0</td>\n",
              "      <td>66.666667</td>\n",
              "      <td>75.0</td>\n",
              "      <td>84.210526</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "                                        title  ... abstractedness_n5\n",
              "0  นักวิจัยออสเตรเลียเผยสาเหตุฉลามโจมตีมนุษย์  ...         84.210526\n",
              "\n",
              "[1 rows x 14 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 10
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "rnCV5X5fNkBf",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 182
        },
        "outputId": "36f4cd30-f9bc-40f6-fcfb-42700b196d09"
      },
      "source": [
        "print('Thai Crosslingual Summarization: Thai to English (TH2EN)')\n",
        "print('Number of Aritcles: ', len(df_th2en))\n",
        "print(\"Column \\'en2th\\' is the back-translation of \\'en_summary\\'\")\n",
        "df_th2en.tail(1)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Thai Crosslingual Summarization: Thai to English (TH2EN)\n",
            "Number of Aritcles:  219336\n",
            "Column 'en2th' is the back-translation of 'en_summary'\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>th_body</th>\n",
              "      <th>th_sum</th>\n",
              "      <th>en_body</th>\n",
              "      <th>en_summary</th>\n",
              "      <th>en2th</th>\n",
              "      <th>url</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>219335</th>\n",
              "      <td>โนวัค ยอโควิช นักเทนนิสมือ 1 ของโลก จากเซอร์เบ...</td>\n",
              "      <td>โนวัค ยอโควิช กลายเป็นนักเทนนิสคนที่ 16 ของปร...</td>\n",
              "      <td>Novak Yeokowich, the world's No.1 tennis playe...</td>\n",
              "      <td>Novak Djokovic became the 16th tennis player i...</td>\n",
              "      <td>โนวัค ดโจโควิค กลายเป็นนักเทนนิสคนที่ 16 ในประ...</td>\n",
              "      <td>https://news.thaipbs.or.th/content/39635</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "                                                  th_body  ...                                       url\n",
              "219335  โนวัค ยอโควิช นักเทนนิสมือ 1 ของโลก จากเซอร์เบ...  ...  https://news.thaipbs.or.th/content/39635\n",
              "\n",
              "[1 rows x 6 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 11
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "5L9LqNNHpByo",
        "colab_type": "text"
      },
      "source": [
        "# Number of Unique Tags and Labels"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "McY8nPIkpfNz",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "#tags\n",
        "unique_tags_trtpbs = set()\n",
        "for index, row in tqdm(df_trtpbs.iterrows(), total=df_trtpbs.shape[0]):\n",
        "  if pd.isnull(row['tags']):\n",
        "    continue\n",
        "  tags = row['tags'].split(',')\n",
        "  unique_tags_trtpbs.update(tags)\n",
        "  \n",
        "unique_tags_tpbs = set()\n",
        "for index, row in tqdm(df_tpbs.iterrows(), total=df_tpbs.shape[0]):\n",
        "  if pd.isnull(row['tags']):\n",
        "    continue\n",
        "  tags = row['tags'].split(',')\n",
        "  unique_tags_tpbs.update(tags)\n",
        "\n",
        "unique_tags_df_tr = set()\n",
        "for index, row in tqdm(df_tr.iterrows(), total=df_tr.shape[0]):\n",
        "  if pd.isnull(row['tags']):\n",
        "    continue\n",
        "  tags = row['tags'].split(',')\n",
        "  unique_tags_df_tr.update(tags)\n",
        "\n",
        "#labels\n",
        "unique_type_trtpbs = set()\n",
        "for index, row in tqdm(df_trtpbs.iterrows(), total=df_trtpbs.shape[0]):\n",
        "  if pd.isnull(row['labels']):\n",
        "    continue\n",
        "  type_ = row['labels'].split(',')\n",
        "  unique_type_trtpbs.update(type_)\n",
        "\n",
        "unique_type_df_tr = set()\n",
        "for index, row in tqdm(df_tr.iterrows(), total=df_tr.shape[0]):\n",
        "  if pd.isnull(row['type']):\n",
        "    continue\n",
        "  type_ = row['type'].split(',')\n",
        "  unique_type_df_tr.update(type_)\n",
        "\n",
        "unique_type_tpbs = set()\n",
        "for index, row in tqdm(df_tpbs.iterrows(), total=df_tpbs.shape[0]):\n",
        "  if pd.isnull(row['label']):\n",
        "    continue\n",
        "  type_ = row['label'].split(',')\n",
        "  unique_type_tpbs.update(type_)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "KymR48ng2oHD",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 348
        },
        "outputId": "3d74da8e-f3d4-47b0-9bfa-e217793b5ea3"
      },
      "source": [
        "fig, (ax1, ax2) = plt.subplots(1, 2, figsize=(12,4))\n",
        "fig.subplots_adjust(wspace=0.3)\n",
        "#tabels\n",
        "objects = ('ThaiPBS-111K','Thairath-222K', 'TR-TPBS')\n",
        "y_pos = np.arange(len(objects))\n",
        "datasets_size = [len(unique_type_tpbs), len(unique_type_df_tr), len(unique_type_trtpbs)]\n",
        "ax1.bar(objects, datasets_size, align='center', alpha=1, width=0.3, color=\"blue\")\n",
        "ax1.set_title('The Number of Unique Labels Comparison')\n",
        "#tags\n",
        "objects2 = ('ThaiPBS-111K','Thairath-222K', 'TR-TPBS')\n",
        "y_pos2 = np.arange(len(objects2))\n",
        "datasets_size2 = [len(unique_tags_tpbs), len(unique_tags_df_tr), len(unique_tags_trtpbs)]\n",
        "ax2.bar(objects2, datasets_size2, align='center', alpha=1, width=0.3, color='pink')\n",
        "ax2.set_title('The Comparison of Number of Unique Tags')\n",
        "plt.show()\n",
        "print()\n",
        "print(\"ThaiPBS-111K   contains\", len(unique_type_tpbs), \"labels and \", len(unique_tags_tpbs), \"tags\")\n",
        "print(\"Thairath-222K  contains\", len(unique_type_df_tr), \"labels and \", len(unique_tags_df_tr), \"tags\")\n",
        "print(\"TR-TPBS        contains\", len(unique_type_trtpbs), \"labels and \", len(unique_tags_trtpbs), \"tags\")"
      ],
      "execution_count": 129,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 864x288 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "text": [
            "\n",
            "ThaiPBS-111K   contains 16 labels and  101456 tags\n",
            "Thairath-222K  contains 40 labels and  445367 tags\n",
            "TR-TPBS        contains 49 labels and  510689 tags\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "XBsYVC4KwADl",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 187
        },
        "outputId": "48272240-3e30-4c5a-f8d9-f4246f0644ce"
      },
      "source": [
        "## Examples of news-labels in TR-TPBS\n",
        "for x in list(unique_type_trtpbs)[:10]:\n",
        "  print(x)"
      ],
      "execution_count": 131,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "ฟุตซอล\n",
            "ข่าวบันเทิง\n",
            "ไลฟ์\n",
            "ไทยพีบีเอส อินไซส์\n",
            "ฟุตบอลโลก\n",
            "เลือกตั้ง\n",
            "ผู้หญิง\n",
            "หวย\n",
            "ต่างประเทศ\n",
            "ทั่วไทย\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ow1MpCNvoF8Q",
        "colab_type": "text"
      },
      "source": [
        "# Average Number of Tokens (words)"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "qTRrEH9PtBQ5",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "trtpbs_title_length = 0\n",
        "trtpbs_article_length = 0\n",
        "trtpbs_summary_length = 0\n",
        "\n",
        "for index, row in tqdm(df_trtpbs.iterrows(), total=df_trtpbs.shape[0]):\n",
        "  trtpbs_title_length += len(word_tokenize(row['title'], keep_whitespace=False))\n",
        "  trtpbs_article_length += len(word_tokenize(row['body'], keep_whitespace=False))\n",
        "  trtpbs_summary_length += len(word_tokenize(row['summary'], keep_whitespace=False))"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "yn2XgSDWvQ0X",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# TR-TPBS\n",
        "avg_trtpbs_title_length = trtpbs_title_length/len(df_trtpbs)\n",
        "avg_trtpbs_article_length = trtpbs_article_length/len(df_trtpbs)\n",
        "avg_trtpbs_summary_length = trtpbs_summary_length/len(df_trtpbs)\n",
        "\n",
        "# Thairath-222k\n",
        "avg_tr_tile_length = df_tr['title_length'].mean()\n",
        "avg_tr_article_length = df_tr['article_length'].mean()\n",
        "avg_tr_summary_length = df_tr['summary_length'].mean()\n",
        "\n",
        "# ThaiPBS-111k\n",
        "avg_tpbs_tile_length = df_tpbs['title_length'].mean()\n",
        "avg_tpbs_article_length = df_tpbs['article_length'].mean()\n",
        "avg_tpbs_summary_length = df_tpbs['summary_length'].mean()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "C4oKOiXMCKiu",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 365
        },
        "outputId": "370cf4bf-f5d3-4400-e18f-c12ab34a7498"
      },
      "source": [
        "# data\n",
        "labels = ['Title',  'Summary', 'Article']\n",
        "tpbs = [ avg_tpbs_tile_length, avg_tpbs_summary_length, avg_tpbs_article_length]\n",
        "tr = [avg_tr_tile_length, avg_tr_summary_length, avg_tr_article_length]\n",
        "tr_tpbs = [avg_trtpbs_title_length, avg_trtpbs_summary_length, avg_trtpbs_article_length]\n",
        "x = np.arange(len(labels))\n",
        "width = 0.22 \n",
        "fig, ax = plt.subplots()\n",
        "rects1 = ax.bar(x-width, tpbs, width, label='ThaiPBS-111k', color='#FFFFFF', hatch=\"//\", edgecolor='black')\n",
        "rects2 = ax.bar(x, tr, width, label='Thairath-222k',color='#ececec', hatch=\".\", edgecolor='black' )\n",
        "rects3 = ax.bar(x+width, tr_tpbs, width, label='TR-TPBS', color='#fde3a7', hatch=\"\", edgecolor='black' )\n",
        "ax.set_ylabel('Average Words')\n",
        "ax.set_title('Average Tokens of each Element')\n",
        "ax.set_xticks(x)\n",
        "ax.set_xticklabels(labels)\n",
        "ax.legend()\n",
        "fig.tight_layout()\n",
        "fig1 = plt.gcf()\n",
        "plt.show()\n",
        "\n",
        "print()\n",
        "print(\"TR-TPBS: avg_title \", avg_trtpbs_title_length,\" avg_body \" ,avg_trtpbs_article_length, \" avg_sum \",avg_trtpbs_summary_length)\n",
        "print(\"Thairath: avg_title\", avg_tr_tile_length, \" avg_body \" ,avg_tr_article_length,  \"  avg_sum \",avg_tr_summary_length )\n",
        "print(\"ThaiPBS: avg_title \", avg_tpbs_tile_length,\" avg_body \" , avg_tpbs_article_length,  \" avg_sum \",avg_tpbs_summary_length )"
      ],
      "execution_count": 134,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "text": [
            "\n",
            "TR-TPBS: avg_title  13.432791082122435  avg_body  452.48069315528454  avg_sum  36.43308375626355\n",
            "Thairath: avg_title 13.870848741648057  avg_body  509.9518856182572   avg_sum  35.91944390774332\n",
            "ThaiPBS: avg_title  11.740516422670991  avg_body  254.28273334886305  avg_sum  36.52586593079721\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "TifXtV0RnSBM",
        "colab_type": "text"
      },
      "source": [
        "# Average Number of Sentences & Compression Ratio"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "64nwIAuawaW8",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# tr-tpbs\n",
        "trtpbs_article_sent_length = 0\n",
        "trtpbs_summary_sent_length = 0\n",
        "for index, row in tqdm(df_trtpbs.iterrows(), total=df_trtpbs.shape[0]):\n",
        "  trtpbs_article_sent_length += len(sent_tokenize(row['body']))\n",
        "  trtpbs_summary_sent_length += len(sent_tokenize(row['summary']))\n",
        "# tpbs\n",
        "tr_article_sent_length = 0\n",
        "tr_summary_sent_length = 0\n",
        "for index, row in tqdm(df_tr.iterrows(), total=df_tr.shape[0]):\n",
        "  tr_article_sent_length += len(sent_tokenize(row['body']))\n",
        "  tr_summary_sent_length += len(sent_tokenize(row['summary']))\n",
        "# tr\n",
        "tpbs_article_sent_length = 0\n",
        "tpbs_summary_sent_length = 0\n",
        "for index, row in tqdm(df_tpbs.iterrows(), total=df_tpbs.shape[0]):\n",
        "  tpbs_article_sent_length += len(sent_tokenize(row['body']))\n",
        "  tpbs_summary_sent_length += len(sent_tokenize(row['summary']))\n",
        "\n",
        "avg_trtpbs_article_sent_length = round(trtpbs_article_sent_length/len(df_tpbs), 2)\n",
        "avg_trtpbs_summary_sent_length = round(trtpbs_summary_sent_length/len(df_tpbs), 2)\n",
        "avg_tr_article_sent_length = round(tr_article_sent_length/len(df_tr), 2)\n",
        "avg_tr_summary_sent_length = round(tr_summary_sent_length/len(df_tr), 2)\n",
        "avg_tpbs_article_sent_length = round(tpbs_article_sent_length/len(df_trtpbs), 2)\n",
        "avg_tpbs_summary_sent_length = round(tpbs_summary_sent_length/len(df_trtpbs), 2)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "pbo4gdybPdxE",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 365
        },
        "outputId": "50622c3b-5e32-4088-fc45-a850407bd489"
      },
      "source": [
        "# data\n",
        "labels = ['Summary', 'Article']\n",
        "tpbs = [avg_tpbs_summary_sent_length, avg_tpbs_article_sent_length]\n",
        "tr = [avg_tr_summary_sent_length, avg_tr_article_sent_length]\n",
        "tr_tpbs = [avg_trtpbs_summary_sent_length, avg_trtpbs_article_sent_length]\n",
        "x = np.arange(len(labels))\n",
        "width = 0.22 \n",
        "fig, ax = plt.subplots()\n",
        "rects1 = ax.bar(x-width, tpbs, width, label='ThaiPBS-111k', color='#FFFFFF', hatch=\"//\", edgecolor='black')\n",
        "rects2 = ax.bar(x, tr, width, label='Thairath-222k',color='#ececec', hatch=\".\", edgecolor='black' )\n",
        "rects3 = ax.bar(x+width, tr_tpbs, width, label='TR-TPBS', color='#fde3a7', hatch=\"\", edgecolor='black' )\n",
        "ax.set_ylabel('Average Sentences')\n",
        "ax.set_title('Average Sentences of Article and Summary')\n",
        "ax.set_xticks(x)\n",
        "ax.set_xticklabels(labels)\n",
        "ax.legend()\n",
        "fig.tight_layout()\n",
        "fig1 = plt.gcf()\n",
        "plt.show()\n",
        "\n",
        "print()\n",
        "print(\"TR-TPBS:  avg_sent_body\", avg_trtpbs_article_sent_length, \" avg_sent_sum\", avg_trtpbs_summary_sent_length, \" compression ratio\", round(avg_trtpbs_article_sent_length/ avg_trtpbs_summary_sent_length, 2))\n",
        "print(\"Thairath: avg_sent_body\", avg_tr_article_sent_length, \" avg_sent_sum\", avg_tr_summary_sent_length, \" compression ratio\",  round(avg_tr_article_sent_length/avg_tr_summary_sent_length, 2))\n",
        "print(\"Tpbs:     avg_sent_body\", avg_tpbs_article_sent_length, \"  avg_sent_sum\", avg_tpbs_summary_sent_length, \"  compression ratio\",  round(avg_tpbs_article_sent_length/avg_tpbs_summary_sent_length, 2))"
      ],
      "execution_count": 141,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAagAAAEYCAYAAAAJeGK1AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nO3deXwV1f3/8deHTUQJQgICskQR6l6VxYpFoNal4oKKuKJAFW1Rq2WpgtJoRVFcSytCKeLSVotW0R+tSyWIC9aCVVCoVpQICApRFPyKCHx+f8y54eaS5QZyk0nyfj4e95FZzsx8ZnLvfO6ZOfeMuTsiIiJxU6+6AxARESmJEpSIiMSSEpSIiMSSEpSIiMSSEpSIiMSSEpSIiMSSElSMmdkFZrbezBpVdyy7yszOMrP5ZjbXzF4xs8fMrFt1xxU3Znalmb0dXpeXUe4PZvaPXdzWIjPbv5wyPczsLTNbvivb2lVmNtbM1phZXhll+pjZy+E9Ns/Mnjaz46swTKlkSlDxdjrQEDixugPZFeEkOBkY4O59gGOBjUCfSlh3rpnVih/zmVlj4C7gJ0TH5otSyjUI848zsxZprntGCSf3Y939g7KWc/c3gKvT2UYmuft44NnS5ptZFvAE8HN37+PuxwJvEn2GpIZSgoopM2sGbAWeAQZWczi76ghgubuvAnD3bcDdwH+rNar42Rto4O6fuPsX7v5YKeV+DNwHrAPO2NmNufv6nV02hr4HbHX3xUnTJgMLqykeqQzurlcMX8BFwACib4BfAbsRfYN04B+hzPnAKuCpMN4VmAe8BLwIHBCmDwOWA48CfwDeAuYCewIPAy8ALxN9oBskxfATYFFY381hHW8BPZJifD3M/zOQVcq+9AC+A04vY39PBOaHdT0DtA3TxwFrgN+FWN8BHgzzmoXte9ifuUB9olrnROA14BXg14ABjUIZB4YDfweWAWckxbFnOEavhNcfgZZhXifg+RDjy0DPpOV+HeLPBx4D2pSyn3sTfdOfF2K/OEzvkLIvt5ZxrKaG9dwNPJ80PXkdg0Os3wK/CMdweVj3T4E7gfXA4KTlLwjLzwmv48L0PkRfMJKP0fRwfF4DLi8j1jOI3ov/DPt8TAmxXhzmv5dyTDuF4zwfeCAc17xSttMG2AL8DKhXwvyHgE1AnzA+K2w7N4z/Lcy/Nsz7EDib6D0+B1gMdC7h8zSF6D35CNAZmAl8QFSTS2y7F/Bc2MfXgP5heur7cTawISzvYb/3BI4P8Syo7vNSlZ8HqzsAvUr5x8CfgMbhTfwF0B9oAHya8iGeRVQTbgasBX4UpvcLH/h6YTwvnKRahvK3AS2AC5PWNQP4aRjOIboMd3QYPw3YlvQBP4boG3zi5D0RmFbG/vw2LL+EKOm0TZq3b/hgfi+MDwf+mRLXf4iSdGOgMCmuXMBTtjU2nFQSyeq1lP10YHQYHgi8lzRvKjA9DNcDniY6QTcAlgJDw7zDwv43BQ4K+2Vh3t2J41TCcfgn4SQbjvFqoFdp+1LC8g2BR8NwIvHnJM3PDft3URj/JdHJewYpJ3eik+PgMNwz8f4I4wOAGWG4D8UT1B/Y/iWhKdHJ84elxHsh0CIpto9LiHVgGB8NPJc0/1/AdWF4H+Dz1H1I2daocDyWA7cTEkrS/OXJ/xeSElTS/N+G4f7AZ4QvL0Tv3/uTyuYBK4g+d42IPpdTib4IHUn0fm4Qyp4M7B+Gs4i+VDZLiWNcGD4vLL8IOD+pzKPAntV1PqquV7UHoFcJ/xTYC5iSND4N+EsYvicxDzgEuCUMXwB8lLKeL9l+Is8DnkyZb0Qn81fCyWo5cF+YdzGwLKX8V2xPUH8AHkiadwTwDeEkXcp+dQFuAQqAr9n+DX0skJ9Urnn40LYJ4zOAu5Lmv5748FJygvofoWYSxq+h+InPgcPC8AHA5jBcj+hb9LFJZTsDrYgS8magYdK8t8MJpT3RyfMsogTSiKSaaFL5fcK2OyRN+wPbE+IO+1LCOk6meK3nA+CypPHcsI1mKcvNoOwEVRRHGG8AHBWG+xASVNIx6p1U9l6S3q8p2ziU6EtU4j3mQKuUWLPC+EnA+2G4Y5jXPmldT6fuQwnbaweMIfrCsIXiX0yWU36COjUM7x/mNw3jl1C8tpoHPJE0/gpwZdKxc7ZfBegA/AV4NRyDTYSrEElxfD9lP0YCz4bhloQvC3Xt1QCJo/7A0WY2N4w3B/Yzs92JLlW8aGZXAYOITjwQfTBbJC0DUY0qO2n8y5TtXAxcBhzu7p+Hm+i5YV4bohpCss+ThtsBByVtL1G7yy5hOQDc/X1gjJldD9xPdFnsxRLWBVES25uohgFRckzYRJQEStMO+KWZDQnjexJdzkqWWN8moqQC0YlgN6Ljloj5fwBm1pfoRPKCmSVm70aUCFaYWT/gV8DviWq/NxCdIFPjInn9YbgirRnPBA40s8FhvCFRLXBKciF3T/1fl6cd0bf2xPJbiGowqRLH6HYz+yZM24vo0m9JngZ+7+53AIQGLU1SYk3+XyT+r23C3+T3UvL7r0TuvpLoS9AtZnZDGH6kvOWSbAh/t4T1JY+nvuc2JA1vSYy7+5bwHkmUfwhY7O7nAYQWkcWOATt+Nv8EjDezNkT/30crsA+1hhJUPJ0IdHX37wBCM/O1QD93f9zMVhFd2/+euy8Ny6wAVnrUSo6wXBbRh740PYA33D3xwW+YNG810ckoWXKLsRXAh+4+PGl7Oe6+Q3Iys6OAA919BkSNJMzsKeDWpHUtcPd+Scs0p3hSqogVwM3uPjOsqx7RSbQ8a4nu2bQkupyHmbUlujS5Avgu5fjuAWwzsybAEnfvb2atie5n/IooAafGRVj/x0nDK9PZqfA+aOLuxyRNOwp41cz2dvdP01lPKVaQ9P8OLQUPdve3U8oljtEV7v7vULYhO55wMbNWRF94nk0ql67EF5PkY5VNVMvZgZntQ1QbHJ80+QmimkjCZqLkipml836oLD2AO5LGyz0O7r7azPKJroz0IboHW+eoFV/MhBPzlkRyAnD3zUQ39BOt+R4iusn9YtKi/w/IMbPuYT17EN2wb1bG5j4Avm9mu4UT0nFJ82YDrczsmLC+04Ddk+bPAPqFeDGz7xE1bijJ7sCwEFMiYZxBdNMcossfR5lZxzC/FdGlkHTenxvCMk3M7Foz+0GI7Xwzqx/KXEx0GbFMHrUufIiogUEizj8CrYlqEx+b2ZlhXgPgKaLLlj2AG8M61hDd+6ufsnrc/ROiBimJ9WcT3St8II39BDiBqNFAsjeI7pWcVc6yG4AmZraHmf2phPkzgJPNLCeMn5OIM1nSMRqUNPl6osYEqQqJaq5HhfGTyokxeTsFRPs2CIoSUO8yFmkIDA3HNGEA299jAB8RXRaH6FJpVfmAcAzM7DC21w7L8xAwguiy59YMxRZv1X2NUa/tL6Jk8hbRTed+SdP7EbU220TUvLgt0f2enJTluxKd2F8i+mCeEqafT/TNcw3wUFL5PYAniWoLM8PwGuCXYf7JRK2X5gLXEV126520/IVsb/X1LNCllP1qSXTv7HWipPkG0Q3lpkllTmD7Nfo5wA/C9F+yvQXaGUQNLNYTNVFPNAj5E7CA6OS/O9HJakLS9h4mqnlA1LLNw7zkVoDPh/l7Et3zS7RQG54UY6ewn4lWfIkGE62Bv4bprxLVoPYq5Vi0Ah6n/FZ8x6cs1zscgzeBfZKmjyVKBGuIGhkkr+OgpHJHh2P2Rng/JFrx/ZfwXgv/z/lh2ceJbuj3IHpPbgJmphyj18I+3wPUL2V/+xO9n58jagmaOPYtkmKdFfY/sZ2Hko73K6HcX8JxXQ4MK2E7TYDxbG9JOZ+U1pTAD4m+PPyT6EtLIpZ92N7K7y2iJuvJsfUIx2k9UeOL5M/Tzyj+njw6xJm8n8cQfcbyiRpbrA7bOYji78cfpezT7kRXEY6o7nNTdb0SrY5EdmBmLXz75T/MbCPQ3bdfVhSRDAk1+Ofcvc72hqFLfFKWp0LvBoRLW58StZATkQwxs6PMrAvRvegXqjue6qRGElKW+cDLobWWA2d51LpLRDKnDdFl65XU8a6adIlPRERiSZf4REQklmrsJb6cnBzPzc2t7jBERGQXLVy4cJ27p/7usuYmqNzcXBYsWFDdYYiIyC4ys4KSpusSn4iIxJISlIiIxJISlIiIxFKNvQdVku+++46VK1eyaVNZ/aNKXDRu3Jh27drRsGFF+hAVkbqiViWolStX0rRpU3Jzc0l6JILEkLtTWFjIypUr2Xfffas7HBGJoVp1iW/Tpk1kZ2crOdUAZkZ2drZquyJSqozUoEInh88QPaKgEVGvxEOJeuedQNS7cWdgjIdn2JjZKKLek5sT9Sz99E5ue5fjl6qh/5WIlCWTl/jmu/vNAGY2i+hJoL2Af7r7X83sVKKHeA0KD13r6+4nh+fsLDWzl7ziTwUtJjc3l4KCEpvX75T27dvz8ccfl19QRER2WUYu8bn7tqTk1IDocdLvET3XaH4o9moYBzglMT10RrqUsh9OlpaCgoJSnzOSn59PTk4O+fn5aT2XJD8/nxUrVpS7zV69ejFy5EiGDBlCs2bNGDlyJCNHjqRVq1ZUtOeLIUOG8J///AeA22+/HTNjzJgx3HDDDZxxxhnMnz+fzz77jMGDB9O2bVvy8vIYNWoUP/7xj/nwww8B2LhxI0OHDmXs2LGMGzeOfv36MWfOnBK39+233zJx4kR23313Nm7cWGzewoUL6d27NyNHjkxrmdGjR9OnT58K7a+ISDGZfNgUUXfxzwI3hvFvCQ9yI6q9efg7Bbg6ablHgEtKWN8wogfTLejQoYOnWrJkSbHxaPd2lJ+f7zk5OZ6fn1/i/NLKl7a+ZNOnT3d398WLF3vHjh2LTU8eT8e2bduKjQO+YcMGd3efM2eOd+/evSi+rl27FpXLy8vza665xt3dJ02aVDTs7v7888/7o48+WuL25s6d68uWLSu2nYQHHnjAb7jhBh8xYkRay3z00Ufeu3fvcvcx9X8mInUPsMBLyCEZbSTh7s+5+0nAvmb2c6JHUzcNs7OALzyqMSVPT8z7rIT1TXX3bu7erWXLHbptSsvcuXM5++yzmTlzZlrf8JPLp2PIkCGlTt+6dSvjxo1jwIABDB06tGjepZdeSl5eHiNHjuTOO+8EYNGiRfTt25cZM2aUuL7Vq1eTk5NT4ry1a9eSOD6tW7fmmWee4YUXXmDLli0cf/zxnHPOOSUu17t3b/bbb78S5w0ePJh69XZ8u5S1TMIVV1xB//79ef7558ssJ1LX5XZsj5nVqFdux/YZOx6ZaiRxELCvu88Okz4C9gNmEz0SeQXRY5AT82cTPTYZM2sIHEj0SOxKtSvJqTIuV61bt45Ro0bRtGlTDjnkEAoLC8nOzuaUU07h9NOjx74cfvjhDBs2jMMOO6zEbd5yyy3Uq1ePr7/+mmnTphVNX7NmDRMmTGD16tW8+eabjB07FoABAwbwzTffcNNNN3Huuedy6qmnMmHCBFq3br3L+5OOl156ibZt2zJp0iQ1ihApR8HHK9m64i/VHUaF1G9/XsbWnalGEt8CPzWzI4BEwrkK2AzcFp4W2QkYCeDur5tZvpndQtSKb4S7r6/MgKo7OQG0atWKpk2jimJOTg4bNmwgOzub1atXM2bMGLKysvjqq68oLCwsKpdqzJgx7LnnnjtMb926Nddeey0A8+fP56STTuLtt98GYNCgQQwaNIjCwkJ++ctfcsUVV/D4449z00038dprr9GpUyd+//vfV8o+JnvvvfeYOHEi++23n5KTiFRYRhKUuy8jarVXkktLWWZiJmKBeCQnKLlZ9dtvv83tt99e1Kjh6ad3qnV9MW3atOGTTz4BIC8vj+OOO45evXqRnZ3NaaedxoMPPgjAuHHjdnlbZenSpQtPP/00xx9/PE8++SRnnHFGRrcnIrVLrepJIlXHjh2LJYW+fftWaPnU8h07dkxruW+++YapU6fy5ZdfMn36dIYOHcq0adP48ssveeKJJ2jRogUFBQVMnz6d6667jgMPPJBLLrmEAw44gFWrVjF9+nQGDhzIvHnzWLx4MccffzyPPfYYELXmu/LKK4vuMa1du5aHH36Y1atXc/PNNwOwZMkSpkyZAkDXrl2ZOHEiL7zwAtu2beP999/ntttuKzHu5cuX88gjjxRt5/zzz+eAAw4A4KGHHmLevHls3ryZqVOnMmzYsDKXmTZtGh9//DEvvvgi/fr1Y+jQoXz22WdcdtllaR9/Eanbauwj37t16+apz4NaunQpBx54YDVFJDtD/zOR7cysRt6D2tU8YmYL3b1b6vRa1dWRiIjUHkpQIiISS0pQIiISS0pQIiISS0pQIiISS7U6QSUeXFhpXXpUsLNXERHZebX6d1AFBQWsW7cOgMLCwqLuf8aPH092dnaZy5ZUvrS+75L16tWLo446isLCQv72t79x6aXR75IfeughmjRpwvLly9OOf8iQIVx11VUcccQRaS8D0Q+N99prLw4//HBWrFjB1VdfzaGHHkpeXl6py7g7F110EV26dGHbtm0sW7aMyZMns8ceezBjxgxef/11OnXqxJtvvsmVV15Jz549WbZsGddffz1HHnkkK1euJDs7m3HjxvHOO+9w1VVXcdFFFzF48OAKxS4iklCrE1SysWPHMmvWLCD6rcHkyZMrtXzC0KFDGTJkCO+88w75+fnccccdABx88MHceOONFYp5+vTpO9VF0Ny5c8nNzeXwww+nffv2nHrqqeUmxm3btrHffvtxww03APCzn/2M+++/nxEjRrBq1SruueceGjduzL/+9S8uueQSFi9ezOeff865555b1I/gQQcdRL9+/ejatSvHHntsheMWEUlWZxJUVSmrN/Nx48Yxbtw4lixZQlZWFtOnTwei3sz32WcfNm7cSJs2bRgxYgSLFi3iqquuYvDgwZx++umcd170Y7gDDjiAOXPm8Jvf/IaFCxeyefNmGjVqxKZNm5g4cSLvv/9+UQ1q+fLlRf3zvfvuu1x55ZW8+eabjBkzhn79+hWLr379+sUS6LZt24r6/EvUJFOnd+/evdg6tm3bxh577FFs2po1azj++OM5+eSTGTFiBK1atdqZwyoidVCdSVDjx48vqo0kugSqzPLp2NnezJs3b861117L6NGjuffee1mzZg1bt27FzIqWO+2003j33Xc5+OCD6dOnD7m5ucUurzVq1IhJkyaxYMECbrzxxh0SVLLly5fz4Ycf8tvf/rbYdHfn3nvv5a677tphmSeffJITTzyxqGukRPmpU6cyZcoUevbsubOHTUTqqDqToLKzs9O+TLcz5dOxq72ZJ7oESjwq47XXXmP06NG0aNGCVatWsXbt2lK3vf/++xfbLlBib+YrV67kuuuu47HHHmO33XYrWt7dGTVqFIMHD+boo48utu78/Hzy8/O55557ik2fPn0669ev58QTT6zQcRIRgTqUoOJgV3szT15+/fr1DBo0iK+++opGjRqxaNGionn169fH3fnkk09o2LBhqdtO7c182bJl5OXlMWXKFLKysnjiiSc466yz2Lp1K9dccw1nnXUWvXv3LpoOMHv2bF5++WXuvfdeVq9eTUFBQVECGzp0KD169GDgwIG88sorNG/ePN1DJSJSuxNUx44d02p5V5H1paOyezPv27cvDz/8MIsWLSpKDs2aNWPgwIFcdNFFdOvWjSVLlvDwww9zzDHH0Lt3b+666y7mzJnDzTffzDPPPMMXX3zBBx98wCOPPEJBQQEvvvgixx13XFHMmzZt4thjj2WfffbhtNNOA6Bz586cddZZjBo1iqeeeqooCS5btoyzzjqLhQsXcs4559CtWzf69u3L119/zfDhw8nOzmbevHk0b96cXr16sddee3HmmWdy66238oMf/KDS/h8iUrupN3OpVvqfiWyn3syLq9U/1BURkZpLCUpERGJJCUpERGJJCUpERGJJCUpERGKpVieo3I7tK7c3847tq3uXRETqjFr9O6iCj1dWapPN+u3PK7dMab2Z33fffbRq1YqePXvSuXNnli1bRufOnfn1r39dbPnBgwcX/Xbrvvvu48ILLyQrK4t//etf3HnnnYwePZrNmzdzwgknsHbtWurVq8e9997L448/zujRo+nZsyedOnVi6dKlHHnkkYwZMwaIuiJ65pln6NChA6tXr2bTpk08+OCDlXZsREQqW61OUNWhrN7MH3zwQc4//3xOOeUUtm3bRk5ODsOGDaNNmzZFy/fu3buow9kZM2YwZswYcnNzeeCBB+jRowd9+vRh48aNRY/O6N27N88++ywDBw7kvvvuK1r/5s2badasGcOHD6dZs2ZceumlfPjhh2RlZQFwySWXVO2BERGpoFp9ia86lNWbebKNGzfSoEEDmjRpslPLA2zZsoV169aV2FvGF198QZMmTYr602vdujXjx49n1apVAEybNq38nRERqUaqQVWxxx9/nFdeeYU5c+bw97//nWbNmlV4Ha+99hp5eXlFD1Xs0aNHsfW//fbbvPDCC1x//fU0btwYgH/84x/ccsstdO/enb333puRI0dywQUXVNp+iYhUNtWgqtiAAQOYMGEC/fr1Y+bMmQDMnz+fk046iZNOOqmo09iy9OzZk7y8PCZNmsT555+/w/rHjh3L3Llzee6553jyyScBaN++PZMnT2bVqlXceeedDB8+nMWLF1f+DoqIVBLVoKrJiBEjOPDAA/nFL37B0UcfzbPPPlvp22jdujWffPIJAMcddxwvvvgiZsaPfvQj2rdvz3fffVfp2xQRqSy1OkF17NAurZZ3FVlfOkrqzfyJJ56goKCAxx57jC5dutClSxeGDx/O0KFDGTdu3A4P9Js6dSrffPMNv/vd77j11ltp2LAhCxYsYN68eWzevLnYIy+AYut/++232bhxI9u2bePiiy8GoEWLFvz85z+nVatWfPLJJwwcOJAjjzyy0o6NiEhly0hv5mbWCbgZeBNoBxS6+01mlgf0SSo63t1fCMuMArKA5sDz7l76g5FQb+a1hf5nItupN/PiMlWDagE86u6zwsaXmNlsAHfvU0JwRwF93f1kM2sALDWzl9z9ywzFJyIiMZeRBOXu/06ZVA/4GsDMxgLfAvWBSe7+f8ApwPyw7BYzWwr0BsqsRYmISO2V8VZ8ZnYG8Jy7/xeYCdzj7ncAG4BJoVirMJ7wVZiWuq5hZrbAzBasXbu2xO3V1Acw1kX6X4lIWTKaoMysL9AXuAbA3d9196/D7DnAj8LwZ0DTpEWzwrRi3H2qu3dz924tW7bcYXuNGzemsLBQJ74awN0pLCws+p2WiEiqjLXiM7N+QC/gF0AbM+sInOnuo0KRzsCyMDwbGBeWawgcCMyr6DbbtWvHypUrKa12JfHSuHFj2rVLr2WkiNQ9GUlQZtYVeAxYAOQDewC/B7aY2b1EtaNDgZ8DuPvrZpZvZrcQteIb4e7rK7rdhg0bsu+++1bSXoiISHXKVCOJhcCeFVxmYiZiERGRmkldHYmISCwpQYmISCwpQYmISCwpQYmISCwpQYmISCwpQYmISCwpQYmISCwpQYmISCwpQYmISCwpQYmISCwpQYmISCwpQYmISCwpQYmISCwpQYmISCwpQYmISCwpQYmISCwpQYmISCwpQYmISCyVm6DMbLiZ/cDMfmRmb5nZ4CqIS0RE6rh0alD7uPvrwPXA+cBBmQ1JREQkvQT1pZm1ATa5+xJgXYZjEhERSStBtQPmA1PM7Bige2ZDEhERgQblFXD3K83sBndfb2bNgMuqIC4REanjyk1QZtYCGGtmW4G5wDLg8wzHJSIidVw6l/gmAIuAbcDrwJUZjUhERIT0EtR77v4g8IW7fw6syHBMIiIiaSWog0MrPg/3oPbLcEwiIiLl34MCHgD+DbQAfg6cm9GIRERESK8V38tAOzPLcXf9BkpERKpEOl0djTWzwe6+zsyGmVmfKohLRETquHQu8WW7+3gAd59qZrcRNTcvlZl1Am4G3iT6oW+hu98UmqxPAD4EOgNj3P3TsMwoIAtoDjzv7k/v3C6JiEhtkE6CWpMyvj6NZVoAj7r7LAAzW2Jms4FLgX+6+1/N7FTgDmCQmR0F9HX3k82sAbDUzF5y9y/T3xUREalN0klQncxsJPAB0AnoWN4C7v7vlEn1gK+BfsD4MO1V4MEwfApRd0q4+xYzWwr0BorVosxsGDAMoEOHDmmELiIiNVU6zcxHENWILgl/R1ZkA2Z2BvCcu/8XaAVsCLO+ApqHGlPy9MS8Vqnrcvep7t7N3bu1bNmyImGIiEgNk04rvo3AmMS4mR1JdG+pXGbWF+gLXB0mfQY0JbpMmEX0498tZpaYnpAVyoqISB2VTl98RwIXsz2BHAZ0S2O5fkAv4BdAGzPrCMwGjibqjeKYME74Oy4s1xA4EJhXkR0REZHaJZ17UJOAu4G1YXxQeQuYWVfgMWABkA/sAfyeqCZ2m5l1IbqfNRLA3V83s3wzu4WoFd8Id0+nMYaIiNRS6SSoRe7+eGLEzP5b3gLuvhDYs5TZl5ayzMQ0YhERkToinQTVyMzGA+8DDpwKnJ3RqEREpM5LpxXfkcBmIBfYl6gln4iISEalU4P6mbu/nhgxs84ZjEdERARIL0G9aWZXAg2BfwH/y2xIIiIi6V3iu4vosl4HYDXw64xGJCIiQnoJqsDdbwRWu/uHwMcZjklERCStBLWfme1G9ETdekD7DMckIiKS1j2o54GPiJqYDwOuyWhEIiIipNcX35Nmlg/sT/ToDT1VV0REMi6dJ+qOdvf17r4AaARMzXxYIiJS15VagzKzDkQ/zj3AzI4Nk+uR3n0rERGRXVLWJb4jgP7A4YCFaVuBZzIdlIiISKkJKjyufZaZdS/hCbkiIiIZVe7lOnf/t5k1N7MO4ZVXBXGJiEgdl84DC/8IHEX0hFsj6lEiL7NhiYhIXZfO76CaufshiREz65O5cERERCLptMh718ySHz7YPFPBiIiIJKSToIYAn5nZR2b2ETAtwzGJiIiklaD+7O5N3H1fd98XGJ3poERERNJpxXetmdUzsxwzM3f/Y1UEJiIidVs6XWccGccAAA9iSURBVB2dAHwITAfOM7PLMh6ViIjUeelc4jsVOAB41d3/DLTNbEgiIiLpJaiV7r6J6HEbAN9mMB4REREgvd9BdTGza4k6jb0C2CfDMYmIiKRVg7oayAJygDbArzIakYiICOk9sHADMKYKYhERESlSag3KzK4ys9vCcHcz+8zMPjezU6ouPBERqavKusR3DHBLGL4pvDoD52U6KBERkbIu8S1x9y9DP3w9gNPc/Tsz+18VxSYiInVYWQkqK/w9E5jr7t+F8d3KW6mZtQZuBr7v7t3DtMHA5cCmUOyP7v5wmHch0RN8twLL3H1KBfdDRERqmbIS1JdmthDYD/gJgJldDxyUxnp/CMwielx8snPdfXnyBDNrB4wEjnB3N7N/m9kcd1dNTUSkDivrke83mdkTwBfu/kmY/DLwl/JW6u6Pl/LcqCvMbA3QBPidu38OnAgsdPfED4HnEyVEJSgRkTqszGbm7v5uyvhLu7Ctl4DZ7r7WzE4GZgLHAa2ADUnlvgrTdmBmw4BhAB06dNiFUEREJO7S+aFupXD3j9x9bRidA/Q2s/pEj5JvmlQ0K0wraR1T3b2bu3dr2bJlZgMWEZFqVWUJysxuNbNEja0zsNzdtwLPAV3NzMK8o4F/VFVcIiIST+n0xYeZ1QNaAIVJ94rKKt8bGAS0CQ0r7gTWAJPDU3kPBS4EcPeVZnYHcLeZbQWmqYGEiIiUm6DC86CmAouAR82saXnNwMO9qtT7VfeWUf4R4JHywxURkbpCz4MSEZFY0vOgREQklvQ8KBERiaWKPg+qNXoelIiIVIF0alDfd3c9D0pERKpUOglqkpn9J2ncgWXApPAwQxERkUqXziW+14FXgYfD30JgBTAxg3GJiEgdl06CWuHuf3T3fHf/I/BleEzGhxmOTURE6rB0ElQPM2sJYGZ7Az8I03MyFpWIiNR56dyDmgwsMrPdgW+AIWbWDViX0chERKROKzdBuftzZtaWqMa0DqgXOnldkOngRESk7kq3N/MDgAOBXsD9mQtHREQkkk5nsbcD3yPqg+99okQlIiKSUenUoL5x99OBx939AuBPGY5JREQkrQTVKPxtHh442DWD8YiIiADpJajvzOxUokYRG4DNmQ1JREQkvWbmC4AF7r7azF5w9y8zHZSIiEg6NajfAFsBlJxEpKZo3bo1ZlajXlJcOgkq390/S4yY2RkZjEdEpFJ8+umnNGzYkLPPPpt169aV+RowYAANGzas9vJSXDqX+A4ws9eBpWH8UODJzIUkIlI5+vfvz80331xuufHjxxfVYOJQXiLm7mUXMPsHMCFp0iB3vySjUaWhW7duvmCBOrMQkZKZWY2rleTk5LB1xV+qO4wKqd/+PMrLI+Uxs4Xu3i11ejo1qPPcfX3SiubvUiQiIiJpSCdB7WVm04mamM8GVgKvZTQqERGp89JpJDEWuBf4CHgKODujEYmIiJBegvqvu78E/J+7bwbWZDgmERGRtBLUYWb2A6CxmR0C7J/hmERERNK6B3UbMA04DDgB+GlGIxIRESG9BJXj7j0zHomIiEiSdC7x3Wlm15pZy4xHIyKSQYWFhVx++eVcfvnlFBYW1rjydU06NairgBXAz80sC/iLu5f5C1kzaw3cDHzf3buHaY2BO4BVQGdggru/H+ZdCBxB1OffMnefspP7IyJSqrFjxzJr1iwg+iHv5MmTY1VeikunBrXC3VcCc4EuwJ/TWOaHwCwg+YhfDXzs7rcCdwN/BDCzdsBIYKS7jwYuMbPOae+BiIjUSunUoB4xs2bAJ8DviX4PVSZ3f9zM+qRM7geMCfMXm9n3Q43sRGChb+8rYz7wE+B/6e2CiEh64tbnXmr5mTNnlrtMXZJOgvoWGODuy8zsSGAS8OOd2FYrot4oEr4K00qbvgMzGwYMA+jQocNOhCAidVl2dna5l93iXL6uSecS3xCgr5n9G/gr0HQnt/VZyrJZYVpp03fg7lPdvZu7d2vZUm02RERqs1ITlJkdYWb3A28BvYG33H1/4KKd3NZs4Oiw7kOBt939K+A5oKttv0N4NPCPndyGiIjUEmXVoOYBewAHufsgok5icff3ylupmfUGBgFtzOx6M9udqD+/jmZ2PTCC8IPf0ADjDuBuM7sTmObuuv8kIlLHlXUPqi1wAXCtmb1MepcDAQh9971UwqzhpZR/BHgk3fWLiEjtV2qCcvcNwP0AZtYD2NPMbgD2dfehVRSfiIjUUem04sPd3wDeCM3C1eREREQyLu3LdgChUcOQDMUiIiJSpEIJCiA8E0pEJPbi0oee+tzbOWld4hMRqYlmzZoVyz739OPc9FS4BiUiIlIVVIMSkVqrf//+sexzT9Jj2/torVm6devmCxaU+dQPEanDzIx169ZVdxgVkpOTw9YVf6nuMCqkfvvz2NU8YmYL3b1b6nRd4hMRkVhSghIRkVhSghIRkVhSghIRkVhSghIRkVhSghIRkVhSghIRkVhSghKROiNufe6pj76yqScJEakz4tbnXmp5KU41KBERiSXVoESkzohbn3up5WfOnFnuMnWJ+uITkVpJffFVDfXFJyIidY4SlIiIxJISlIiIxJISlIiIxJISlIiIxJISlIiIxJISlIiIxJISlIjUWnHpQ0997u0c9SQhIrXWrFmzYtnnXnnlJaIalIiIxFK11KDM7HVgUxjd6u7HmVkLYALwIdAZGOPun1ZHfCJSO/Tv3z+Wfe5JeqqlLz4zy3P3vJRp9wNz3P2vZnYqMNDdB5W2DvXFJyJlUV98VaM29sV3qJn9yszyzKxfmNYPmB+GXw3jxZjZMDNbYGYL1q5dW1WxiohINaiuRhK3ufsbZlYfmGdmG4BWwIYw/yuguZk1cPctiYXcfSowFaIaVFUHLSIiVadaalDu/kb4uxV4GegLfAY0DUWygC+Sk5OIiNQtVZ6gzOwAM/tp0qTOwDJgNnB0mHZMGBcRkTqqOi7xfQX0M7O2RDWlFcCfgb8Dt5lZF6ATMLIaYhMRkZio8gTl7p8AZ5Yw63Pg0ioOR0REYko/1BURkVhSghKROiNufe6pj76yqS8+Eakz4tbnXmp5KU41KBERiSXVoESkzohbn3up5WfOnFnuMnVJtfTFVxnUF5+IlEV98VWN2tgXn4iISJmUoEREJJaUoEREJJaUoEREJJaUoEREJJaUoCTjcju2x8xqzCu3Y/vqPmQign4HJVWg4OOVNarpbP3251V3CCKCalAiUovFpQ899bm3c1SDEpFaa9asWbHsc6+88hJRDUpERGJJNSgRqbX69+8fyz73JD3qi08yzsxqXCOJmvq5yKTc3FwKCgqqO4wKUV98mae++ESk2hUUFODuRa/8/HxycnLIz88vNr20V1WXl5pPCUpEKmzu3LmcffbZzJw5kz59+tS48lIzKEGJSIXELdkoOdVeSlAikra4JRslp9pNCUpE0hanZKPkVPspQYlI2uKSbJSc6gb9DqqGqYlNfaX2iEOyUXKqO5SgapiCgoJiv+0oLCxk7NixQPRjwOzs7DKXr47yOTk5Ze+U1BpxT05x/7xIcUpQNVzc+hBTn2N1V9yTE8Tv/Z9aXorTPSgR2WVxS05z584tt4zEn2pQNVzc+hBTn2N1TxyT09lnn13ivLi9/1PLz5w5s9xl6pJY9cVnZj8GzgQ+A9zdbyytbF3ti8/M1L9YhqkvvpKZ2Q7HJa7JaebMmfTt21eflSqQyb74YlODMrMmwP3Awe7+rZk9YWbHufuLmdqmWsSJ7Lw4Jye17qsd4nQP6migwN2/DeOvAv0yucHkzi/j3vFl4iUSB3FLNkpOtVNsLvGZ2XnAOe7eP4xfAvRx9wuTygwDhoXR7wHvVXmgtVcOULOuh4hUH31eKldHd2+ZOjE2l/iI7js1TRrPCtOKuPtUYGpVBlVXmNmCkq4Bi8iO9HmpGnG6xDcf6Ghmu4XxY4DZ1RiPiIhUo9jUoNz9/8zsZ8BvzWwtsCiTDSRERCTeYpOgANz9BeCF6o6jjtKlU5H06fNSBWLTSEJERCRZnO5BiYiIFFGCEhGRWFKCqmHM7HIzm2xmeWb2kJlNqO6YRGoqM8sys41m1rWccoPNbK+k8efMrFUZ5S8wsy8qM9a6SPegahAzywKWAa3c3c2sAfA7d7+8mkMTqZFCy+EfAhvd/bIyys0FBrv78jBuXs7J08yWu3tu5UVb98SqFZ+U61vAgF+a2YPuvg643MwuB65191wzO4GohVEfoCHwAPAR8AXQA7iLqFup7sCt7j7bzEYBvwZGAD8GvgHmAn2BFsBp7r7VzH4DNAI2A43dfZSZdQ7bWA2sAU4Fpod1DQL+CfwVeMfdr83kwRHZCZ2Ba4AlZjYCOBSYDCwk+rwNAK4AcoGrzey/wBKin8Nc7e5zzewYYDDwPtHn6g53fyN5IyERfo+o94lmwOjyEpxAWv3C6RWfF3Aw8DCwFngFOClMX55UZi6QG4YHAw+H4f7Ay2H4COCZpGWWA53C8FvAcWF4FtA1DJ+eVP5poo59E9v4axjuBOwFvAR0D9PuI9TW9dIrLi/gKOCMMPwIMCwM5wG3h+HDiL7oFX2mwvQZRF8CDVgJ7B2mHwIcFYaXh78HAkvZfsVqRvJnSa/SX6pB1TDu/i4wyMzqEz2a5G9m1r6cxZaFv+uThr+geNdSuHt55RqZ2e3A58A+QHLfWUuT12FmvwV+Ee6RLfHwyRSJkXOAr83s+0S1pcvY/vumxPt5EZT5tNscIMvdPw3l3ymhzCHANuBXYT3fEXXlJuVQgqpBzCwXuMHdf+rRJbcngSklFC0vYe3MtvciqrlluftmMzsspUhqAnoKuBUYDQyv7HhEdoWZNQPWu/tNSdOWJTWWSH0/b42K2CGE5BWsA740s1bu/ln4XOzp7q8llVkMfOPuE8J2jiRKUlIOJaia5Usg28zuDsP7Ar9y90Ize9jMJgHvAhuI7k3dTXRPqLmZdSG6J3RY+ICcRtT34fFEl+SamdlgoADoCAw2s6eJLnEMIuor8a/AQ2a2ADiIqCa3NGkbRc/vCgn0D0Bbd99QBcdGJC2hv8/JRPdSE9MOILoyMDFMOtTM3nH3xFNRnwWuJbrcN53tn4t/A+cCt5rZ/4C2wPVmdgHRZ+pyd7/fzKaY2V1El+bbAtdlej9rA7Xik0pnZo1CLWsk8Dd3/7C6YxKRmkc1KMmEC82sG/CpkpOI7CzVoEREJJbUk4SIiMSSEpSIiMSSEpSIiMSSEpSIiMSSEpSIiMTS/wfSJ+RUWAKRWQAAAABJRU5ErkJggg==\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "text": [
            "\n",
            "TR-TPBS:  avg_sent_body 298.74  avg_sent_sum 27.18  compression ratio 10.99\n",
            "Thairath: avg_sent_body 120.53  avg_sent_sum 10.13  compression ratio 11.9\n",
            "Tpbs:     avg_sent_body 20.47   avg_sent_sum 2.96   compression ratio 6.92\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ylHFbAfsbSzO",
        "colab_type": "text"
      },
      "source": [
        "# Level of Abstractedness"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Az8pduD_br2w",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 348
        },
        "outputId": "f38125a2-d448-4188-a911-fb4b5a926bff"
      },
      "source": [
        "# data\n",
        "labels = ['1-grams', '2-grams', '3-grams', '4-grams', '5-grams', 'Sentences']\n",
        "trtpbs = [24.53102012,56.69915252,69.68104127,76.32939739,80.20300963,99.82606719]\n",
        "tr = [23.66051313,55.44594339,68.05996836,74.38170649,78.0331832,95.09129881] \n",
        "tpbs = [23.79801516,58.54822143,73.93684608,82.50715153,87.75524999,98.46708886]\n",
        "x = np.arange(len(labels)) \n",
        "width = 0.22  \n",
        "fig, ax = plt.subplots()\n",
        "rects1 = ax.bar(x-width, tpbs, width, label='ThaiPBS-111K', color='#FFFFFF', hatch=\"//\", edgecolor='black')\n",
        "rects2 = ax.bar(x, tr, width, label='Thairath-222K',color='#ececec', hatch=\".\", edgecolor='black' )\n",
        "rects3 = ax.bar(x+width, trtpbs , width, label='TR-TPBS', color='#fde3a7', hatch=\"\", edgecolor='black' )\n",
        "ax.set_ylabel('Proportion of novel n-gram')\n",
        "ax.set_title('Abstractedness Level in each Dataset')\n",
        "ax.set_xticks(x)\n",
        "ax.set_xticklabels(labels)\n",
        "ax.legend()\n",
        "fig.tight_layout()\n",
        "fig1 = plt.gcf()\n",
        "plt.show()\n",
        "\n",
        "print()\n",
        "print(\"\\\"The higher level of abstractedness creates new challenges for the summarization systems \\n as they have to be more creative in generating more novel summaries\\\", according to Koupaee, M., & Wang, W. Y. (2018).\")"
      ],
      "execution_count": 150,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "text": [
            "\n",
            "\"The higher level of abstractedness creates new challenges for the summarization systems \n",
            " as they have to be more creative in generating more novel summaries\", according to Koupaee, M., & Wang, W. Y. (2018).\n"
          ],
          "name": "stdout"
        }
      ]
    }
  ]
}