graph_prototype_v1.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "authorship_tag": "ABX9TyPgaEHB8Lp0KKjE1bCE1HIU",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/ErsinDemirel94/docs2graph/blob/main/graph_prototype_v1.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Input Files"
      ],
      "metadata": {
        "id": "kmTlTY1W_RsX"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "Aşağıda liste şeklinde verilen paragraflar işlenerek, isimli varlıklar tespit edilir(named entity recognition). Aynı varlığı ifade eden tespitler düzeltilir. (deduplication)İki farklı varlığın aynı paragrafta bahsi geçiyorsa aralarında bağlantı tanımlanır. (co-occurrence graph)"
      ],
      "metadata": {
        "id": "tMPhsqb73ZZ1"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "with open('dummy_text.txt') as f:\n",
        "    lines = f.readlines()\n",
        "\n",
        "lines = [i.replace(\"\\n\",\"\") for i in lines] \n",
        "list_of_paragraphs = [ele for ele in lines if len(ele) > 1]\n",
        "list_of_paragraphs"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Ey5pAJvEN39H",
        "outputId": "822ea5d5-a605-46cf-b529-3a1c4e1b21c7"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[\"Bu ilanda Mehmet Demir'in ortağı olduğu Vestel A.Ş'ye dair bilgiler paylaşılacaktır. Bu firmanın diğer, ortakları ise sırasıyla Ayşe Yeşil ve Fatma Kara'dır.\",\n",
              " \"Bu diğer ilanda ise 1995 yılında kurulan Borok Anonim Şirketi'nin konkardatosu ilan edilmiştir. Bu durumda Fatma Kara ve Halis Uçak işsiz kalacaktır.\",\n",
              " \"1994'daki karar üzerine Karpuzlu AVM'nin üst kuruluşu Yetişen Holding'in yönetim kurulu başkanı Halis Uçak, Antalya'ya yerleşmiştir.\",\n",
              " \"Yetişen Holdingin bir diğer önemli kişisi Kaan Gök, Mollalar LTD. ŞTİ'nin sahibidir.\",\n",
              " 'Geçenlerde Kaan gok bir açıklama yaptı. Bu açıklama da Borok firmasında çalışan Fatma Kara ile ilgiliydi. Ve çok önemliydi.']"
            ]
          },
          "metadata": {},
          "execution_count": 55
        }
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "D9Vymeez_Olf",
        "outputId": "66079a15-4d56-418c-cce7-364adc764859"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[{'ilan_id': 0, 'ilan_text': \"Bu ilanda Mehmet Demir'in ortağı olduğu Vestel A.Ş'ye dair bilgiler paylaşılacaktır. Bu firmanın diğer, ortakları ise sırasıyla Ayşe Yeşil ve Fatma Kara'dır.\"}, {'ilan_id': 1, 'ilan_text': \"Bu diğer ilanda ise 1995 yılında kurulan Borok Anonim Şirketi'nin konkardatosu ilan edilmiştir. Bu durumda Fatma Kara ve Halis Uçak işsiz kalacaktır.\"}, {'ilan_id': 2, 'ilan_text': \"1994'daki karar üzerine Karpuzlu AVM'nin üst kuruluşu Yetişen Holding'in yönetim kurulu başkanı Halis Uçak, Antalya'ya yerleşmiştir.\"}, {'ilan_id': 3, 'ilan_text': \"Yetişen Holdingin bir diğer önemli kişisi Kaan Gök, Mollalar LTD. ŞTİ'nin sahibidir.\"}, {'ilan_id': 4, 'ilan_text': 'Geçenlerde Kaan gok bir açıklama yaptı. Bu açıklama da Borok firmasında çalışan Fatma Kara ile ilgiliydi. Ve çok önemliydi.'}]\n"
          ]
        }
      ],
      "source": [
        "import json\n",
        "\n",
        "# list_of_paragraphs\n",
        "ilans = []\n",
        "for idx, paragraphs in enumerate(list_of_paragraphs):\n",
        "  dct_ = {\"ilan_id\":idx,\n",
        "  \"ilan_text\":paragraphs}\n",
        "  ilans.append(dct_)\n",
        "\n",
        "#j$sonString = json.dumps(output)\n",
        "\n",
        "print(ilans)\n",
        "\n",
        "\n",
        "# ilan_id, ilan_text in ilans"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "N tane JSON Record with :\n",
        "\n",
        "\n",
        "*   id: str\n",
        "\n",
        "*   text: str\n",
        "\n",
        "\n"
      ],
      "metadata": {
        "id": "ucehsq_W_fyn"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "# NER Pipeline"
      ],
      "metadata": {
        "id": "AvfojuA4_18i"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "!pip install transformers==4.4.2"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "4Z1xIOmr_6nJ",
        "outputId": "b1de48cc-1a72-4fcc-cadb-bf7f7c41b65b"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Installing collected packages: tokenizers, sacremoses, transformers\n",
            "Successfully installed sacremoses-0.0.53 tokenizers-0.10.3 transformers-4.4.2\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "!pip install https://huggingface.co/turkish-nlp-suite/tr_core_news_trf/resolve/main/tr_core_news_trf-any-py3-none-any.whl"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Deh2xpoE_4eI",
        "outputId": "41136733-9713-47d6-dcbd-72aa7a8b57a9"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Installing collected packages: spacy-alignments, spacy-transformers, tr-core-news-trf\n",
            "Successfully installed spacy-alignments-0.8.6 spacy-transformers-1.1.8 tr-core-news-trf-3.4.2\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import spacy\n",
        "\n",
        "class nlp_pipeline:\n",
        "\n",
        "  def __init__(self, gpu=True, disable_list = [], batch_size = 16, lang_model = \"tr_core_news_trf\"):\n",
        "\n",
        "    self.gpu = gpu\n",
        "    self.disable_list = disable_list\n",
        "    self.batch_size = batch_size\n",
        "    self.lang_model = lang_model\n",
        "\n",
        "\n",
        "    if gpu == True:\n",
        "      #spacy.require_gpu(gpu_id=0) # add GPU \n",
        "      spacy.prefer_gpu(gpu_id=0) # add GPU \n",
        "\n",
        "\n",
        "    self.nlp = spacy.load(self.lang_model, disable= self.disable_list)\n",
        "\n",
        "    # ['transformer','tagger','morphologizer','trainable_lemmatizer','parser','ner']\n",
        "\n",
        "\n",
        "  def pipe_process(self, list_of_texts, batch_size):\n",
        "\n",
        "    docs = []\n",
        "    \n",
        "    ## NLP pipe list olarak alır ve batch process yapar. n_process ile parallel de yapar.\n",
        "    for doc in self.nlp.pipe(list_of_texts, batch_size = self.batch_size): \n",
        "      docs.append(doc)\n",
        "\n",
        "    return docs\n",
        "\n",
        "  def ner_pipe(self, list_of_texts=[], batch_size = 16):\n",
        "    \n",
        "\n",
        "    # ner classes: CARDINAL, DATE, EVENT, FAC, GPE, LANGUAGE, LAW, LOC, MONEY, \n",
        "    # NORP, ORDINAL, ORG, PER, PERCENT, PERSON, PRODUCT, QUANTITY, TIME, TITLE, WORK_OF_ART\n",
        "\n",
        "    # PARAGRAPH İŞLEYEBİLİR. in ilan, convert list of paragraphs. or convert list of sentences, apply ner.\n",
        "\n",
        "    docs = []\n",
        "    \n",
        "     ## NLP pipe list olarak alır ve batch process yapar. n_process ile parallel de yapar.\n",
        "    for idx, doc in enumerate(self.nlp.pipe(list_of_texts, batch_size = batch_size)):\n",
        "      docs.append((doc,idx)) \n",
        "\n",
        "    out = []\n",
        "    ent_label_list = [\"ORG\", \"PERSON\"]\n",
        "    \n",
        "    for doc, idx in docs:\n",
        "\n",
        "      for ent in doc.ents:\n",
        "        # simple preprocessing of entities\n",
        "        entity_name = ent.text.split(\"\\'\", 1)[0]\n",
        "\n",
        "        if ent.label_ in ent_label_list:\n",
        "          out.append((entity_name, ent.label_))\n",
        "\n",
        "    return out\n",
        "\n"
      ],
      "metadata": {
        "id": "Gfozofes_UTS"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "!pip install zemberek-python"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "w9U486mKR2k1",
        "outputId": "c6cbd5c4-7fef-4635-f248-a8596d4be4aa"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Looking in indexes: https://pypi.org/simple, https://us-python.pkg.dev/colab-wheels/public/simple/\n",
            "Collecting zemberek-python\n",
            "  Downloading zemberek_python-0.2.2-py3-none-any.whl (95.0 MB)\n",
            "\u001b[K     |████████████████████████████████| 95.0 MB 103 kB/s \n",
            "\u001b[?25hCollecting antlr4-python3-runtime==4.8\n",
            "  Downloading antlr4-python3-runtime-4.8.tar.gz (112 kB)\n",
            "\u001b[K     |████████████████████████████████| 112 kB 48.4 MB/s \n",
            "\u001b[?25hRequirement already satisfied: numpy>=1.19.0 in /usr/local/lib/python3.7/dist-packages (from zemberek-python) (1.21.6)\n",
            "Building wheels for collected packages: antlr4-python3-runtime\n",
            "  Building wheel for antlr4-python3-runtime (setup.py) ... \u001b[?25l\u001b[?25hdone\n",
            "  Created wheel for antlr4-python3-runtime: filename=antlr4_python3_runtime-4.8-py3-none-any.whl size=141231 sha256=7c6d41e82639f0a5c4d4108a4938cb9773f49544d47cbadb82ffe42800d1f493\n",
            "  Stored in directory: /root/.cache/pip/wheels/ca/33/b7/336836125fc9bb4ceaa4376d8abca10ca8bc84ddc824baea6c\n",
            "Successfully built antlr4-python3-runtime\n",
            "Installing collected packages: antlr4-python3-runtime, zemberek-python\n",
            "Successfully installed antlr4-python3-runtime-4.8 zemberek-python-0.2.2\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Tokenization & Deduplication"
      ],
      "metadata": {
        "id": "XySEGCz5xhZG"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import time\n",
        "import logging\n",
        "\n",
        "from zemberek import (\n",
        "    TurkishSpellChecker,\n",
        "    TurkishSentenceNormalizer,\n",
        "    TurkishSentenceExtractor,\n",
        "    TurkishMorphology,\n",
        "    TurkishTokenizer\n",
        ")\n",
        "\n",
        "extractor = TurkishSentenceExtractor()"
      ],
      "metadata": {
        "id": "mBSY5NHgR1Ap"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "!pip install pandas_dedupe"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "vhY3MUmoSvfS",
        "outputId": "19fcdc58-9a97-4920-a044-7cbb3770ce32"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Installing collected packages: unidecode, pandas-dedupe\n",
            "Successfully installed pandas-dedupe-1.5.0 unidecode-1.3.6\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import pandas as pd\n",
        "import pandas_dedupe\n",
        "\n",
        "nlp_parser = nlp_pipeline(gpu=True,batch_size=4)"
      ],
      "metadata": {
        "id": "WlwGy_kJT8T5"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "ilans"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "kv5_NE-LURJs",
        "outputId": "2dec01fc-392d-4109-d6f9-9320c031c402"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[{'ilan_id': 0,\n",
              "  'ilan_text': \"Bu ilanda Mehmet Demir'in ortağı olduğu Vestel A.Ş'ye dair bilgiler paylaşılacaktır. Bu firmanın diğer, ortakları ise sırasıyla Ayşe Yeşil ve Fatma Kara'dır.\"},\n",
              " {'ilan_id': 1,\n",
              "  'ilan_text': \"Bu diğer ilanda ise 1995 yılında kurulan Borok Anonim Şirketi'nin konkardatosu ilan edilmiştir. Bu durumda Fatma Kara ve Halis Uçak işsiz kalacaktır.\"},\n",
              " {'ilan_id': 2,\n",
              "  'ilan_text': \"1994'daki karar üzerine Karpuzlu AVM'nin üst kuruluşu Yetişen Holding'in yönetim kurulu başkanı Halis Uçak, Antalya'ya yerleşmiştir.\"},\n",
              " {'ilan_id': 3,\n",
              "  'ilan_text': \"Yetişen Holdingin bir diğer önemli kişisi Kaan Gök, Mollalar LTD. ŞTİ'nin sahibidir.\"},\n",
              " {'ilan_id': 4,\n",
              "  'ilan_text': 'Geçenlerde Kaan gok bir açıklama yaptı. Bu açıklama da Borok firmasında çalışan Fatma Kara ile ilgiliydi. Ve çok önemliydi.'}]"
            ]
          },
          "metadata": {},
          "execution_count": 74
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "final = []\n",
        "for ilan in ilans:\n",
        "  ilan_text =ilan[\"ilan_text\"]\n",
        "  ilan_id = ilan[\"ilan_id\"]\n",
        "\n",
        "  list_of_sents = extractor.from_paragraph(ilan_text)\n",
        "  list_of_tups = nlp_parser.ner_pipe(list_of_texts= list_of_sents)\n",
        "\n",
        "  for name,type_ in list_of_tups:\n",
        "    final.append((name, type_, ilan_id))\n",
        "\n",
        "# CONVERT TO DATAFRAME  \n",
        "\n",
        "df = pd.DataFrame(final, columns =['name', 'type', 'ilan_id'])"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "R8AtcucTUBiB",
        "outputId": "7e223e7b-a7fe-4838-ef56-d3439c46dad4"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/torch/amp/autocast_mode.py:198: UserWarning: User provided device_type of 'cuda', but CUDA is not available. Disabling\n",
            "  warnings.warn('User provided device_type of \\'cuda\\', but CUDA is not available. Disabling')\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "df.head(20)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 520
        },
        "id": "-GUj9g3VWXBD",
        "outputId": "e0d94c99-0d18-4944-c9a3-31fe55ca48f3"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "                    name    type ilan_id\n",
              "0           Mehmet Demir  PERSON       0\n",
              "1             Vestel A.Ş     ORG       0\n",
              "2             Ayşe Yeşil  PERSON       0\n",
              "3             Fatma Kara  PERSON       0\n",
              "4   Borok Anonim Şirketi     ORG       1\n",
              "5             Fatma Kara  PERSON       1\n",
              "6             Halis Uçak  PERSON       1\n",
              "7        Yetişen Holding     ORG       2\n",
              "8             Halis Uçak  PERSON       2\n",
              "9      Yetişen Holdingin     ORG       3\n",
              "10              Kaan Gök  PERSON       3\n",
              "11     Mollalar LTD. ŞTİ     ORG       3\n",
              "12              Kaan gok  PERSON       4\n",
              "13                 Borok     ORG       4\n",
              "14            Fatma Kara  PERSON       4"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-802ba738-b7cd-41d8-8e29-7f5af1e4a304\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <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>name</th>\n",
              "      <th>type</th>\n",
              "      <th>ilan_id</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>Mehmet Demir</td>\n",
              "      <td>PERSON</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>Vestel A.Ş</td>\n",
              "      <td>ORG</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>Ayşe Yeşil</td>\n",
              "      <td>PERSON</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>Fatma Kara</td>\n",
              "      <td>PERSON</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>Borok Anonim Şirketi</td>\n",
              "      <td>ORG</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>Fatma Kara</td>\n",
              "      <td>PERSON</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6</th>\n",
              "      <td>Halis Uçak</td>\n",
              "      <td>PERSON</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7</th>\n",
              "      <td>Yetişen Holding</td>\n",
              "      <td>ORG</td>\n",
              "      <td>2</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>8</th>\n",
              "      <td>Halis Uçak</td>\n",
              "      <td>PERSON</td>\n",
              "      <td>2</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9</th>\n",
              "      <td>Yetişen Holdingin</td>\n",
              "      <td>ORG</td>\n",
              "      <td>3</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>10</th>\n",
              "      <td>Kaan Gök</td>\n",
              "      <td>PERSON</td>\n",
              "      <td>3</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>11</th>\n",
              "      <td>Mollalar LTD. ŞTİ</td>\n",
              "      <td>ORG</td>\n",
              "      <td>3</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>12</th>\n",
              "      <td>Kaan gok</td>\n",
              "      <td>PERSON</td>\n",
              "      <td>4</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>13</th>\n",
              "      <td>Borok</td>\n",
              "      <td>ORG</td>\n",
              "      <td>4</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>14</th>\n",
              "      <td>Fatma Kara</td>\n",
              "      <td>PERSON</td>\n",
              "      <td>4</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-802ba738-b7cd-41d8-8e29-7f5af1e4a304')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-802ba738-b7cd-41d8-8e29-7f5af1e4a304 button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-802ba738-b7cd-41d8-8e29-7f5af1e4a304');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 90
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# DEDUPLICATE DF\n",
        "deduplicated_df = pandas_dedupe.dedupe_dataframe(df,[\"name\", \"type\"])# update_model=True\n",
        "\n",
        "# 5 ten fazla örnek işaretlemek gerekiyor.\n",
        "# Sonrasında json output olarak kaydediyor training i\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "id": "7QOgvac-CWLC",
        "outputId": "9594ae9d-4a1d-49b4-8fa0-c7c1b3d3b382"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/pandas_dedupe/utility_functions.py:17: FutureWarning: The default value of regex will change from True to False in a future version.\n",
            "  df[i] = df[i].str.replace('[^\\w\\s\\.\\-\\(\\)\\,\\:\\/\\\\\\\\]','')\n",
            "name : yetisen holding\n",
            "type : org\n",
            "\n",
            "name : yetisen holdingin\n",
            "type : org\n",
            "\n",
            "0/10 positive, 0/10 negative\n",
            "Do these records refer to the same thing?\n",
            "(y)es / (n)o / (u)nsure / (f)inished\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Importing data ...\n",
            "Starting active labeling...\n",
            "y\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "name : borok anonim sirketi\n",
            "type : org\n",
            "\n",
            "name : borok\n",
            "type : org\n",
            "\n",
            "1/10 positive, 0/10 negative\n",
            "Do these records refer to the same thing?\n",
            "(y)es / (n)o / (u)nsure / (f)inished / (p)revious\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "y\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "name : mollalar ltd. sti\n",
            "type : org\n",
            "\n",
            "name : borok\n",
            "type : org\n",
            "\n",
            "2/10 positive, 0/10 negative\n",
            "Do these records refer to the same thing?\n",
            "(y)es / (n)o / (u)nsure / (f)inished / (p)revious\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "n\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "name : yetisen holdingin\n",
            "type : org\n",
            "\n",
            "name : borok\n",
            "type : org\n",
            "\n",
            "2/10 positive, 1/10 negative\n",
            "Do these records refer to the same thing?\n",
            "(y)es / (n)o / (u)nsure / (f)inished / (p)revious\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "n\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "name : yetisen holding\n",
            "type : org\n",
            "\n",
            "name : borok\n",
            "type : org\n",
            "\n",
            "2/10 positive, 2/10 negative\n",
            "Do these records refer to the same thing?\n",
            "(y)es / (n)o / (u)nsure / (f)inished / (p)revious\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "n\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "name : halis ucak\n",
            "type : person\n",
            "\n",
            "name : halis ucak\n",
            "type : person\n",
            "\n",
            "2/10 positive, 3/10 negative\n",
            "Do these records refer to the same thing?\n",
            "(y)es / (n)o / (u)nsure / (f)inished / (p)revious\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "y\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "name : kaan gok\n",
            "type : person\n",
            "\n",
            "name : kaan gok\n",
            "type : person\n",
            "\n",
            "3/10 positive, 3/10 negative\n",
            "Do these records refer to the same thing?\n",
            "(y)es / (n)o / (u)nsure / (f)inished / (p)revious\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "y\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "name : fatma kara\n",
            "type : person\n",
            "\n",
            "name : fatma kara\n",
            "type : person\n",
            "\n",
            "4/10 positive, 3/10 negative\n",
            "Do these records refer to the same thing?\n",
            "(y)es / (n)o / (u)nsure / (f)inished / (p)revious\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "y\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "name : fatma kara\n",
            "type : person\n",
            "\n",
            "name : fatma kara\n",
            "type : person\n",
            "\n",
            "5/10 positive, 3/10 negative\n",
            "Do these records refer to the same thing?\n",
            "(y)es / (n)o / (u)nsure / (f)inished / (p)revious\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "y\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "name : fatma kara\n",
            "type : person\n",
            "\n",
            "name : fatma kara\n",
            "type : person\n",
            "\n",
            "6/10 positive, 3/10 negative\n",
            "Do these records refer to the same thing?\n",
            "(y)es / (n)o / (u)nsure / (f)inished / (p)revious\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "y\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "name : vestel a.s\n",
            "type : org\n",
            "\n",
            "name : mollalar ltd. sti\n",
            "type : org\n",
            "\n",
            "7/10 positive, 3/10 negative\n",
            "Do these records refer to the same thing?\n",
            "(y)es / (n)o / (u)nsure / (f)inished / (p)revious\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "n\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "name : vestel a.s\n",
            "type : org\n",
            "\n",
            "name : yetisen holdingin\n",
            "type : org\n",
            "\n",
            "7/10 positive, 4/10 negative\n",
            "Do these records refer to the same thing?\n",
            "(y)es / (n)o / (u)nsure / (f)inished / (p)revious\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "n\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "name : vestel a.s\n",
            "type : org\n",
            "\n",
            "name : borok anonim sirketi\n",
            "type : org\n",
            "\n",
            "7/10 positive, 5/10 negative\n",
            "Do these records refer to the same thing?\n",
            "(y)es / (n)o / (u)nsure / (f)inished / (p)revious\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "n\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "name : vestel a.s\n",
            "type : org\n",
            "\n",
            "name : borok\n",
            "type : org\n",
            "\n",
            "7/10 positive, 6/10 negative\n",
            "Do these records refer to the same thing?\n",
            "(y)es / (n)o / (u)nsure / (f)inished / (p)revious\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "n\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "name : vestel a.s\n",
            "type : org\n",
            "\n",
            "name : yetisen holding\n",
            "type : org\n",
            "\n",
            "7/10 positive, 7/10 negative\n",
            "Do these records refer to the same thing?\n",
            "(y)es / (n)o / (u)nsure / (f)inished / (p)revious\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "(y)es / (n)o / (u)nsure / (f)inished / (p)revious\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "n\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "name : halis ucak\n",
            "type : person\n",
            "\n",
            "name : kaan gok\n",
            "type : person\n",
            "\n",
            "7/10 positive, 8/10 negative\n",
            "Do these records refer to the same thing?\n",
            "(y)es / (n)o / (u)nsure / (f)inished / (p)revious\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "n\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "name : halis ucak\n",
            "type : person\n",
            "\n",
            "name : kaan gok\n",
            "type : person\n",
            "\n",
            "7/10 positive, 9/10 negative\n",
            "Do these records refer to the same thing?\n",
            "(y)es / (n)o / (u)nsure / (f)inished / (p)revious\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "n\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "name : halis ucak\n",
            "type : person\n",
            "\n",
            "name : kaan gok\n",
            "type : person\n",
            "\n",
            "7/10 positive, 10/10 negative\n",
            "Do these records refer to the same thing?\n",
            "(y)es / (n)o / (u)nsure / (f)inished / (p)revious\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "n\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "name : halis ucak\n",
            "type : person\n",
            "\n",
            "name : kaan gok\n",
            "type : person\n",
            "\n",
            "7/10 positive, 11/10 negative\n",
            "Do these records refer to the same thing?\n",
            "(y)es / (n)o / (u)nsure / (f)inished / (p)revious\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "n\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "name : borok anonim sirketi\n",
            "type : org\n",
            "\n",
            "name : mollalar ltd. sti\n",
            "type : org\n",
            "\n",
            "7/10 positive, 12/10 negative\n",
            "Do these records refer to the same thing?\n",
            "(y)es / (n)o / (u)nsure / (f)inished / (p)revious\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "n\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "name : fatma kara\n",
            "type : person\n",
            "\n",
            "name : halis ucak\n",
            "type : person\n",
            "\n",
            "7/10 positive, 13/10 negative\n",
            "Do these records refer to the same thing?\n",
            "(y)es / (n)o / (u)nsure / (f)inished / (p)revious\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "(y)es / (n)o / (u)nsure / (f)inished / (p)revious\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "f\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Finished labeling\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Clustering...\n",
            "# duplicate sets 9\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "                    name    type ilan_id  cluster id  confidence\n",
              "0           mehmet demir  person       0           5    1.000000\n",
              "1             vestel a.s     org       0           6    1.000000\n",
              "2             ayse yesil  person       0           7    1.000000\n",
              "3             fatma kara  person       0           0    0.683446\n",
              "4   borok anonim sirketi     org       1           1    0.640009\n",
              "5             fatma kara  person       1           0    0.683446\n",
              "6             halis ucak  person       1           2    0.683446\n",
              "7        yetisen holding     org       2           3    0.673818\n",
              "8             halis ucak  person       2           2    0.683446\n",
              "9      yetisen holdingin     org       3           3    0.673818\n",
              "10              kaan gok  person       3           4    0.683446\n",
              "11     mollalar ltd. sti     org       3           8    1.000000\n",
              "12              kaan gok  person       4           4    0.683446\n",
              "13                 borok     org       4           1    0.640009\n",
              "14            fatma kara  person       4           0    0.683446"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-6ad9213d-c437-4f3c-919f-a0c426f3b4a5\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <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>name</th>\n",
              "      <th>type</th>\n",
              "      <th>ilan_id</th>\n",
              "      <th>cluster id</th>\n",
              "      <th>confidence</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>mehmet demir</td>\n",
              "      <td>person</td>\n",
              "      <td>0</td>\n",
              "      <td>5</td>\n",
              "      <td>1.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>vestel a.s</td>\n",
              "      <td>org</td>\n",
              "      <td>0</td>\n",
              "      <td>6</td>\n",
              "      <td>1.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>ayse yesil</td>\n",
              "      <td>person</td>\n",
              "      <td>0</td>\n",
              "      <td>7</td>\n",
              "      <td>1.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>fatma kara</td>\n",
              "      <td>person</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0.683446</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>borok anonim sirketi</td>\n",
              "      <td>org</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>0.640009</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>fatma kara</td>\n",
              "      <td>person</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0.683446</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6</th>\n",
              "      <td>halis ucak</td>\n",
              "      <td>person</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>0.683446</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7</th>\n",
              "      <td>yetisen holding</td>\n",
              "      <td>org</td>\n",
              "      <td>2</td>\n",
              "      <td>3</td>\n",
              "      <td>0.673818</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>8</th>\n",
              "      <td>halis ucak</td>\n",
              "      <td>person</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>0.683446</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9</th>\n",
              "      <td>yetisen holdingin</td>\n",
              "      <td>org</td>\n",
              "      <td>3</td>\n",
              "      <td>3</td>\n",
              "      <td>0.673818</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>10</th>\n",
              "      <td>kaan gok</td>\n",
              "      <td>person</td>\n",
              "      <td>3</td>\n",
              "      <td>4</td>\n",
              "      <td>0.683446</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>11</th>\n",
              "      <td>mollalar ltd. sti</td>\n",
              "      <td>org</td>\n",
              "      <td>3</td>\n",
              "      <td>8</td>\n",
              "      <td>1.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>12</th>\n",
              "      <td>kaan gok</td>\n",
              "      <td>person</td>\n",
              "      <td>4</td>\n",
              "      <td>4</td>\n",
              "      <td>0.683446</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>13</th>\n",
              "      <td>borok</td>\n",
              "      <td>org</td>\n",
              "      <td>4</td>\n",
              "      <td>1</td>\n",
              "      <td>0.640009</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>14</th>\n",
              "      <td>fatma kara</td>\n",
              "      <td>person</td>\n",
              "      <td>4</td>\n",
              "      <td>0</td>\n",
              "      <td>0.683446</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-6ad9213d-c437-4f3c-919f-a0c426f3b4a5')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-6ad9213d-c437-4f3c-919f-a0c426f3b4a5 button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-6ad9213d-c437-4f3c-919f-a0c426f3b4a5');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 91
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "ilan_id_lst = deduplicated_df.groupby('cluster id')['ilan_id'].apply(list)\n",
        "ilan_type_lst = deduplicated_df.groupby('cluster id')['type'].apply(list)\n",
        "ilan_aliases_lst = deduplicated_df.groupby('cluster id')['name'].apply(list)\n",
        "\n",
        "# Resolve Type by Majority Vote\n",
        "# ????? \n",
        "\n",
        "fin = pd.concat([ilan_id_lst,ilan_aliases_lst, ilan_type_lst], axis=1)\n",
        "fin[\"entity_ids\"] = fin.index\n",
        "fin = fin.rename(columns={'cluster id': 'entity_id', 'name': 'aliases'})\n",
        "\n",
        "fin = fin[['entity_ids', 'type', 'aliases', 'ilan_id']]\n",
        "\n",
        "# SAVE NODES JSON\n",
        "#fin.to_json('nodes.json', orient='records', lines=True)\n",
        "\n",
        "\n",
        "fin.head()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 238
        },
        "id": "vN5Yix_mDieC",
        "outputId": "d2435f81-a9af-4b0d-d02f-1db34c594569"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "            entity_ids                      type  \\\n",
              "cluster id                                         \n",
              "0                    0  [person, person, person]   \n",
              "1                    1                [org, org]   \n",
              "2                    2          [person, person]   \n",
              "3                    3                [org, org]   \n",
              "4                    4          [person, person]   \n",
              "\n",
              "                                         aliases    ilan_id  \n",
              "cluster id                                                   \n",
              "0           [fatma kara, fatma kara, fatma kara]  [0, 1, 4]  \n",
              "1                  [borok anonim sirketi, borok]     [1, 4]  \n",
              "2                       [halis ucak, halis ucak]     [1, 2]  \n",
              "3           [yetisen holding, yetisen holdingin]     [2, 3]  \n",
              "4                           [kaan gok, kaan gok]     [3, 4]  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-3fee7b60-a632-4525-b896-ad90ff10e2a0\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <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>entity_ids</th>\n",
              "      <th>type</th>\n",
              "      <th>aliases</th>\n",
              "      <th>ilan_id</th>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>cluster id</th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>0</td>\n",
              "      <td>[person, person, person]</td>\n",
              "      <td>[fatma kara, fatma kara, fatma kara]</td>\n",
              "      <td>[0, 1, 4]</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>1</td>\n",
              "      <td>[org, org]</td>\n",
              "      <td>[borok anonim sirketi, borok]</td>\n",
              "      <td>[1, 4]</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>2</td>\n",
              "      <td>[person, person]</td>\n",
              "      <td>[halis ucak, halis ucak]</td>\n",
              "      <td>[1, 2]</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>3</td>\n",
              "      <td>[org, org]</td>\n",
              "      <td>[yetisen holding, yetisen holdingin]</td>\n",
              "      <td>[2, 3]</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>4</td>\n",
              "      <td>[person, person]</td>\n",
              "      <td>[kaan gok, kaan gok]</td>\n",
              "      <td>[3, 4]</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-3fee7b60-a632-4525-b896-ad90ff10e2a0')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-3fee7b60-a632-4525-b896-ad90ff10e2a0 button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-3fee7b60-a632-4525-b896-ad90ff10e2a0');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 102
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "ent_ids = fin.entity_ids.tolist()\n",
        "alias = fin.aliases.tolist()\n",
        "sel_alias = [i[0] for i in alias]\n",
        "\n",
        "id2name = dict(zip(ent_ids,sel_alias))"
      ],
      "metadata": {
        "id": "OzSfQEzjacdw"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "id2name"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "DjstBVN2bCwK",
        "outputId": "e210c4cc-2cc9-4880-a9b7-0647c1276a6c"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "{0: 'fatma kara',\n",
              " 1: 'borok anonim sirketi',\n",
              " 2: 'halis ucak',\n",
              " 3: 'yetisen holding',\n",
              " 4: 'kaan gok',\n",
              " 5: 'mehmet demir',\n",
              " 6: 'vestel a.s',\n",
              " 7: 'ayse yesil',\n",
              " 8: 'mollalar ltd. sti'}"
            ]
          },
          "metadata": {},
          "execution_count": 113
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "sel_alias"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "1FtHlHw2an2K",
        "outputId": "36780d58-316d-49cc-d2d3-36e739b5c9f6"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "['fatma kara',\n",
              " 'borok anonim sirketi',\n",
              " 'halis ucak',\n",
              " 'yetisen holding',\n",
              " 'kaan gok',\n",
              " 'mehmet demir',\n",
              " 'vestel a.s',\n",
              " 'ayse yesil',\n",
              " 'mollalar ltd. sti']"
            ]
          },
          "metadata": {},
          "execution_count": 109
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Create Relations"
      ],
      "metadata": {
        "id": "JnYrdSWexxa3"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "ent_ids = fin.entity_ids.tolist()\n",
        "ilan_ids = fin.ilan_id.tolist()\n",
        "\n",
        "\n",
        "zipped = zip(ent_ids, ilan_ids)"
      ],
      "metadata": {
        "id": "53a0KqwLD7D5"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "Output = {}\n",
        "for x, y in ent_mentions:\n",
        "    if y in Output:\n",
        "        Output[y].append((x))\n",
        "    else:\n",
        "        Output[y] = [(x)]\n",
        "  \n",
        "# Her dokumanda geçen entityler\n",
        "print(Output)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "z6C21xwwD9oX",
        "outputId": "ef9ac6a0-825c-4a2a-fdf1-3aeed12ddbfd"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "{'0': [0, 5, 6, 7], '1': [0, 1, 2], '4': [0, 1, 4], '2': [2, 3], '3': [3, 4, 8]}\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import itertools\n",
        "fin = []\n",
        "for item in Output.items():\n",
        "\n",
        "  doc_id = item[0]\n",
        "  entity_list = item[1]\n",
        "  edges = itertools.combinations(entity_list, 2)\n",
        "  edges = list(set(edges))\n",
        "  for u,v in list(edges):\n",
        "    if u != v:\n",
        "      fin.append((u,v,doc_id))"
      ],
      "metadata": {
        "id": "8bV1eZRWEAq1"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "list_of_relations = []\n",
        "for u,v,r in fin:\n",
        "  list_of_relations.append((u,v))\n",
        "\n",
        "\n",
        "print(len(list_of_relations))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ejeM3kZ5EEi2",
        "outputId": "e1d74478-5c3b-4ff2-8749-6a5c56b89727"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "16\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Visualization"
      ],
      "metadata": {
        "id": "cU3myLUux0Wz"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import networkx as nx\n",
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "\n",
        "from matplotlib.pyplot import figure\n",
        "\n",
        "figure(figsize=(12, 8), dpi=80)\n",
        "\n",
        "\n",
        "G = nx.Graph()\n",
        "G.add_edges_from(list_of_relations)\n",
        "\n",
        "\n",
        "nx.draw(G, cmap = plt.get_cmap('jet'), labels= id2name, with_labels = True)\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 673
        },
        "id": "kvtwTBeyEGqx",
        "outputId": "9cacc527-fc49-420c-f116-535bfc091d68"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 960x640 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "MwzBliZLbfgH"
      },
      "execution_count": null,
      "outputs": []
    }
  ]
}