lectures/aula_named_entity_recognition_crf.ipynb

{
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/viniciusrpb/cic0269_natural_language_processing/blob/main/lectures/aula_named_entity_recognition_crf.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "bwHBSHrMVlJL"
      },
      "source": [
        "# Capítulo 14: Reconhecimento de Entidades Nomeadas\n",
        "\n",
        "\n",
        "## 14.2. Conditional Random Fields\n",
        "\n",
        "https://github.com/howl-anderson/addons/blob/add_crf_tutorial/docs/tutorials/layers_crf.ipynb\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Fcwyky08UmsZ"
      },
      "source": [
        "![The-BiLSTM-CRF-model-CS-confidence-score-BiLSTM-bidirectional-long-short-term-memory.png](data:image/png;base64,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)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "zGjqkmd8Upmy"
      },
      "source": [
        "Fonte da Imagem: Batbaatar, E., & Ryu, K. H. (2019). Ontology-based healthcare named entity recognition from twitter messages using a recurrent neural network approach. International journal of environmental research and public health, 16(19), 3628."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "3xzqTxT27ZZL"
      },
      "outputs": [],
      "source": [
        "#!pip install -U tensorflow-datasets\n",
        "#!pip install keras\n",
        "#!pip install tensorflow\n",
        "!pip install tensorflow-addons[tensorflow]"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "YurnZIGf8fkS"
      },
      "outputs": [],
      "source": [
        "import matplotlib.pyplot as plt\n",
        "import numpy as np\n",
        "import tensorflow as tf\n",
        "import tensorflow_datasets as tfds\n",
        "import tensorflow_addons as tfa\n",
        "import pandas as pd\n",
        "from tensorflow.keras.optimizers import SGD\n",
        "from tensorflow.keras.preprocessing.text import Tokenizer\n",
        "from tensorflow.keras.preprocessing.sequence import pad_sequences\n",
        "from keras.models import Sequential\n",
        "from keras.layers import Dense,Activation,Embedding,Dropout,LSTM,Bidirectional,TimeDistributed,Input,Reshape\n",
        "from keras.utils.np_utils import to_categorical\n",
        "from sklearn.metrics import classification_report\n",
        "from tensorflow_addons.text.crf_wrapper import CRFModelWrapper\n",
        "from tensorflow_addons.layers.crf import CRF\n",
        "from collections import Counter\n",
        "import tensorflow_addons as tfa"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "wbSPBoYBausW"
      },
      "outputs": [],
      "source": [
        "def preprocessDataFrame(df):\n",
        "\n",
        "    dic = {}\n",
        "    dic['tokens'] = []\n",
        "\n",
        "    for text in df['tokens']:\n",
        "        tokens = []\n",
        "        for x in text:\n",
        "            tokens.append(x.decode('utf-8'))\n",
        "        l = \" \".join(tokens)\n",
        "        dic['tokens'].append(l.split())\n",
        "        \n",
        "    res_df = pd.DataFrame.from_dict(dic)\n",
        "    res_df['ner'] = df['ner']\n",
        "    return res_df"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "ZQMmPm1A8j3C"
      },
      "outputs": [],
      "source": [
        "ds_train = tfds.load('conll2003', split='train', shuffle_files=True)\n",
        "ds_valid = tfds.load('conll2003', split='dev', shuffle_files=False)\n",
        "ds_test = tfds.load('conll2003', split='test', shuffle_files=False)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "PVBs34YrVR6t"
      },
      "outputs": [],
      "source": [
        "df_train = preprocessDataFrame(tfds.as_dataframe(ds_train))\n",
        "df_valid = preprocessDataFrame(tfds.as_dataframe(ds_valid))\n",
        "df_test = preprocessDataFrame(tfds.as_dataframe(ds_test))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 424
        },
        "id": "gJ77VPYStB6J",
        "outputId": "0a70ce16-4d3f-4ed7-c350-11bfd0d2b10f"
      },
      "outputs": [
        {
          "data": {
            "text/html": [
              "\n",
              "  <div id=\"df-cc43a93b-b202-4c65-9e5e-f601802a6585\">\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>tokens</th>\n",
              "      <th>ner</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>[\", If, they, 're, saying, at, least, 20, perc...</td>\n",
              "      <td>[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, ...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>[Lauck, 's, lawyer, vowed, he, would, appeal, ...</td>\n",
              "      <td>[1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, ...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>[Thailand, 's, powerful, military, thinks, the...</td>\n",
              "      <td>[5, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 2, 0, ...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>[A, forensic, scientist, who, examined, the, s...</td>\n",
              "      <td>[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, ...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>[Werder, Bremen, 3, 0, 1, 2, 4, 6, 1]</td>\n",
              "      <td>[3, 4, 0, 0, 0, 0, 0, 0, 0]</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>14037</th>\n",
              "      <td>[\", He, was, not, involved, ...]</td>\n",
              "      <td>[0, 0, 0, 0, 0, 0]</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>14038</th>\n",
              "      <td>[\", It, goes, without, saying, that, we, 're, ...</td>\n",
              "      <td>[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, ...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>14039</th>\n",
              "      <td>[Bowling, :]</td>\n",
              "      <td>[0, 0]</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>14040</th>\n",
              "      <td>[National, League]</td>\n",
              "      <td>[7, 8]</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>14041</th>\n",
              "      <td>[LOME, 1996-08-25]</td>\n",
              "      <td>[5, 0]</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>14042 rows × 2 columns</p>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-cc43a93b-b202-4c65-9e5e-f601802a6585')\"\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-cc43a93b-b202-4c65-9e5e-f601802a6585 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-cc43a93b-b202-4c65-9e5e-f601802a6585');\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",
              "  "
            ],
            "text/plain": [
              "                                                  tokens  \\\n",
              "0      [\", If, they, 're, saying, at, least, 20, perc...   \n",
              "1      [Lauck, 's, lawyer, vowed, he, would, appeal, ...   \n",
              "2      [Thailand, 's, powerful, military, thinks, the...   \n",
              "3      [A, forensic, scientist, who, examined, the, s...   \n",
              "4                  [Werder, Bremen, 3, 0, 1, 2, 4, 6, 1]   \n",
              "...                                                  ...   \n",
              "14037                   [\", He, was, not, involved, ...]   \n",
              "14038  [\", It, goes, without, saying, that, we, 're, ...   \n",
              "14039                                       [Bowling, :]   \n",
              "14040                                 [National, League]   \n",
              "14041                                 [LOME, 1996-08-25]   \n",
              "\n",
              "                                                     ner  \n",
              "0      [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, ...  \n",
              "1      [1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, ...  \n",
              "2      [5, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 2, 0, ...  \n",
              "3      [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, ...  \n",
              "4                            [3, 4, 0, 0, 0, 0, 0, 0, 0]  \n",
              "...                                                  ...  \n",
              "14037                                 [0, 0, 0, 0, 0, 0]  \n",
              "14038  [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, ...  \n",
              "14039                                             [0, 0]  \n",
              "14040                                             [7, 8]  \n",
              "14041                                             [5, 0]  \n",
              "\n",
              "[14042 rows x 2 columns]"
            ]
          },
          "execution_count": 139,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "df_train"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "i5qwk5AwWGiu"
      },
      "outputs": [],
      "source": [
        "def label2int():\n",
        "    iob_labels = [\"B\", \"I\"]\n",
        "    ner_labels = [\"PER\", \"ORG\", \"LOC\", \"MISC\"]\n",
        "    all_labels = [(label1, label2) for label2 in ner_labels for label1 in iob_labels]\n",
        "    all_labels = [\"-\".join([a, b]) for a, b in all_labels]\n",
        "    dic = dict(zip(range(1, len(all_labels) + 1), all_labels))\n",
        "    dic[0] = 'O'\n",
        "    return dic "
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "wburyOkU0tsa",
        "outputId": "0abb8098-cf1b-4a6c-b95e-82d63d09ae5e"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "{1: 'B-PER',\n",
              " 2: 'I-PER',\n",
              " 3: 'B-ORG',\n",
              " 4: 'I-ORG',\n",
              " 5: 'B-LOC',\n",
              " 6: 'I-LOC',\n",
              " 7: 'B-MISC',\n",
              " 8: 'I-MISC',\n",
              " 0: 'O'}"
            ]
          },
          "execution_count": 141,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "int2tag = label2int()\n",
        "\n",
        "int2tag"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "NKGag80BWVX_",
        "outputId": "25fa13b7-115c-4cee-8ec3-9f13422c7ae0"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "{'B-PER': 1, 'I-PER': 2, 'B-ORG': 3, 'I-ORG': 4, 'B-LOC': 5, 'I-LOC': 6, 'B-MISC': 7, 'I-MISC': 8, 'O': 0}\n"
          ]
        }
      ],
      "source": [
        "tag2int = {}\n",
        "for key in int2tag:\n",
        "    value = int2tag[key]\n",
        "    tag2int[value] = key\n",
        "print(tag2int)\n",
        "\n",
        "num_labels = len(tag2int)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "e7euVuZV9_Ri"
      },
      "source": [
        "Cria o vocabulário"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "h8ydPLuy9_an"
      },
      "outputs": [],
      "source": [
        "word2index = {}\n",
        "word2index['<OOV>'] = 0\n",
        "word2index['<BEG>'] = 1\n",
        "word2index['<END>'] = 2\n",
        "\n",
        "ind = 3\n",
        "for text in df_train['tokens']:\n",
        "    for word in text:\n",
        "        if word not in word2index:\n",
        "            word2index[word] = ind\n",
        "            ind+=1\n",
        "\n",
        "index2word = {}\n",
        "for key in word2index:\n",
        "    value = word2index[key]\n",
        "    index2word[value] = key"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "RMv8s9mC89n3"
      },
      "outputs": [],
      "source": [
        "def rebuild_datasets(df,labels):\n",
        "\n",
        "    data = []\n",
        "    novel_labels = []\n",
        "\n",
        "    idx = 0\n",
        "    for text in df:\n",
        "\n",
        "        data.append(['<BEG>'])\n",
        "        novel_labels.append(0)\n",
        "\n",
        "        for i in range(0,len(text)):\n",
        "            lista = ['<BEG>']+text[:i+1]\n",
        "            data.append(lista)\n",
        "            if i < len(text)-1:\n",
        "                novel_labels.append(labels[idx][i])\n",
        "            else:\n",
        "                novel_labels.append(0)\n",
        "        idx+=1\n",
        "\n",
        "        data.append(lista+['<END>'])\n",
        "        novel_labels.append(0)\n",
        "\n",
        "    return data,novel_labels"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "wf9UeaQM8-yw"
      },
      "outputs": [],
      "source": [
        "x_train,y_train_int = rebuild_datasets(df_train['tokens'],df_train['ner'])\n",
        "x_valid,y_valid_int = rebuild_datasets(df_valid['tokens'],df_valid['ner'])\n",
        "x_test,y_test_int = rebuild_datasets(df_test['tokens'],df_test['ner'])"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "pF8zNwDQeUOd"
      },
      "source": [
        "Verificando se está tudo certo"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "PbEnKVm2_Zhm",
        "outputId": "51f763cc-d48e-445b-c38c-e2a9c62267be"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Word: <BEG> text label: O int label: 0 \n",
            "Word: \" text label: O int label: 0 \n",
            "Word: If text label: O int label: 0 \n",
            "Word: they text label: O int label: 0 \n",
            "Word: 're text label: O int label: 0 \n",
            "Word: saying text label: O int label: 0 \n",
            "Word: at text label: O int label: 0 \n",
            "Word: least text label: O int label: 0 \n",
            "Word: 20 text label: O int label: 0 \n",
            "Word: percent text label: O int label: 0 \n",
            "Word: , text label: O int label: 0 \n",
            "Word: then text label: O int label: 0 \n",
            "Word: their text label: O int label: 0 \n",
            "Word: internal text label: O int label: 0 \n",
            "Word: forecasts text label: O int label: 0 \n",
            "Word: are text label: O int label: 0 \n",
            "Word: probably text label: O int label: 0 \n",
            "Word: saying text label: O int label: 0 \n",
            "Word: 25 text label: O int label: 0 \n",
            "Word: or text label: O int label: 0 \n",
            "Word: 30 text label: O int label: 0 \n",
            "Word: percent text label: O int label: 0 \n",
            "Word: , text label: O int label: 0 \n",
            "Word: \" text label: O int label: 0 \n",
            "Word: said text label: O int label: 0 \n",
            "Word: one text label: O int label: 0 \n",
            "Word: Sydney text label: B-LOC int label: 5 \n",
            "Word: media text label: O int label: 0 \n",
            "Word: analyst text label: O int label: 0 \n",
            "Word: who text label: O int label: 0 \n",
            "Word: declined text label: O int label: 0 \n",
            "Word: to text label: O int label: 0 \n",
            "Word: be text label: O int label: 0 \n",
            "Word: named text label: O int label: 0 \n"
          ]
        }
      ],
      "source": [
        "for i,word in enumerate(x_train[33]):\n",
        "    print(f'Word: {word} text label: {int2tag[y_train_int[i]]} int label: {y_train_int[i]} ')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "E40GuN4S--9N",
        "outputId": "8b817fb2-d745-411a-e1a9-5146052c7bbc"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "115"
            ]
          },
          "execution_count": 147,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "max_length = max([len(sentence) for sentence in x_train])\n",
        "max_length"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "gN6nP0bMAGPz"
      },
      "outputs": [],
      "source": [
        "def text2sequences(data,vocab):\n",
        "    corpus_int = []\n",
        "    for instance in data:\n",
        "        instance_int = []\n",
        "        for term in instance:\n",
        "            if term in vocab:\n",
        "                instance_int.append(vocab[term])\n",
        "            else:\n",
        "                instance_int.append(vocab['<OOV>'])\n",
        "        corpus_int.append(instance_int)\n",
        "    \n",
        "    return corpus_int"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "2gcb70fIAAvO"
      },
      "outputs": [],
      "source": [
        "train_sequences = text2sequences(x_train,word2index)\n",
        "valid_sequences = text2sequences(x_valid,word2index)\n",
        "test_sequences = text2sequences(x_test,word2index)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "AwDZSyDDCB8Q"
      },
      "outputs": [],
      "source": [
        "trunc_type = 'post'\n",
        "padding_type = 'pre'\n",
        "\n",
        "train_padded = pad_sequences(train_sequences, maxlen=max_length, padding=padding_type, truncating=trunc_type)\n",
        "valid_padded = pad_sequences(valid_sequences, maxlen=max_length, padding=padding_type, truncating=trunc_type)\n",
        "test_padded = pad_sequences(test_sequences, maxlen=max_length, padding=padding_type, truncating=trunc_type)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "y7dUqlViCLL3"
      },
      "outputs": [],
      "source": [
        "y_train = to_categorical(y_train_int)\n",
        "#y_train = y_train.reshape((y_train.shape[0], y_train.shape[1], 1))\n",
        "#y_valid = y_valid.reshape((y_valid.shape[0], y_valid.shape[1], 1))\n",
        "#y_test = y_test.reshape((y_test.shape[0], y_test.shape[1], 1))\n",
        "#y_train = np.expand_dims(y_train,2)\n",
        "y_valid = to_categorical(y_valid_int)\n",
        "#y_valid = np.expand_dims(y_valid,2)\n",
        "y_test = to_categorical(y_test_int)\n",
        "#y_test = np.expand_dims(y_test,2)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "948xJCRTCnae"
      },
      "outputs": [],
      "source": [
        "vocab_size = len(word2index)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "HXktiN4VEUmR"
      },
      "outputs": [],
      "source": [
        "backbone = Sequential()\n",
        "backbone.add(Embedding(input_dim=vocab_size+3,output_dim=64,input_length=max_length))\n",
        "backbone.add(Bidirectional(LSTM(256,activation='relu')))\n",
        "backbone.add(Dropout(0.5))\n",
        "backbone.add(Dense(num_labels,activation='softmax'))\n",
        "backbone.add(Reshape(target_shape=(num_labels,1)))\n",
        "\n",
        "model = CRFModelWrapper(backbone, num_labels)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "background_save": true,
          "base_uri": "https://localhost:8080/"
        },
        "id": "iQyNvJ2pEjuZ",
        "outputId": "70658b0f-7d67-4a17-9a55-1049e87d34b1"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Epoch 1/10\n",
            "3621/3621 [==============================] - 4487s 1s/step - loss: 3.2673 - crf_loss: 3.2673 - val_loss: 1.0748 - val_crf_loss: 1.0748\n",
            "Epoch 2/10\n",
            "3621/3621 [==============================] - 4527s 1s/step - loss: 1.4663 - crf_loss: 1.4663 - val_loss: 0.5967 - val_crf_loss: 0.5967\n",
            "Epoch 3/10\n",
            "3621/3621 [==============================] - 4450s 1s/step - loss: 1.2867 - crf_loss: 1.2867 - val_loss: 0.4564 - val_crf_loss: 0.4564\n",
            "Epoch 4/10\n",
            " 822/3621 [=====>........................] - ETA: 53:04 - loss: 1.2340 - crf_loss: 1.2340"
          ]
        }
      ],
      "source": [
        "f1scoretfa = tfa.metrics.F1Score(num_classes=num_labels, average='macro',threshold=0.5)\n",
        "\n",
        "sgd = SGD(learning_rate = 0.00005)\n",
        "model.compile(optimizer=sgd)\n",
        "#model.compile(optimizer=\"rmsprop\")\n",
        "history_fine = model.fit(train_padded,y_train,batch_size=64,validation_data=(valid_padded,y_valid),epochs=10)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "kQInP_lcFyMU"
      },
      "outputs": [],
      "source": [
        "plt.plot(history_fine.history['loss'])\n",
        "plt.plot(history_fine.history['val_loss'])\n",
        "plt.title('model loss')\n",
        "plt.ylabel('loss')\n",
        "plt.xlabel('epoch')\n",
        "plt.legend(['training set','validation set'], loc='upper left')\n",
        "plt.show()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "NgUE31ItF267"
      },
      "outputs": [],
      "source": [
        "plt.plot(history_fine.history['f1_score'])\n",
        "plt.plot(history_fine.history['val_f1_score'])\n",
        "plt.title('model f1 score')\n",
        "plt.ylabel('f1score')\n",
        "plt.xlabel('epoch')\n",
        "plt.legend(['training set','validation set'], loc='upper left')\n",
        "plt.show()"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "F6Vneb89F4WU"
      },
      "source": [
        "Teste"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "VeyDxErCF49M"
      },
      "outputs": [],
      "source": [
        "y_prob = model.predict(test_padded)\n",
        "y_pred = np.argmax(y_prob,axis=1)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "uOkDZ9JrF6H7"
      },
      "outputs": [],
      "source": [
        "print(classification_report(y_test_int,y_pred))"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Mogb-FZ0F-JD"
      },
      "source": [
        "Deployment"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "ZI3t_a1bF_Rz"
      },
      "outputs": [],
      "source": [
        "sentenca_teste = [\"Mayara is living in Spain which its land urges for coffee\".split()]\n",
        "sentenca_int = text2sequences(sentenca_teste,word2index)\n",
        "sentenca_padded = pad_sequences(sentenca_int, maxlen=max_length, padding=padding_type, truncating=trunc_type)\n",
        "\n",
        "y_prob = model.predict(sentenca_padded)\n",
        "y_pred = np.argmax(y_prob,axis=1)\n",
        "\n",
        "for idx,word in enumerate(sentenca_teste):\n",
        "    print(f'{word} is label {int2tag[y_pred[idx]]}')"
      ]
    }
  ],
  "metadata": {
    "accelerator": "GPU",
    "colab": {
      "provenance": [],
      "authorship_tag": "ABX9TyNuJ5GocENIpflfgaXZLp9R",
      "include_colab_link": true
    },
    "kernelspec": {
      "display_name": "Python 3",
      "name": "python3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "nbformat": 4,
  "nbformat_minor": 0
}