Копия_блокнота_"5_nlp_libraries_ipynb".ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "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/shevtsovasofi/programming_sofi/blob/main/%D0%9A%D0%BE%D0%BF%D0%B8%D1%8F_%D0%B1%D0%BB%D0%BE%D0%BA%D0%BD%D0%BE%D1%82%D0%B0_%225_nlp_libraries_ipynb%22.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Подготовка\n",
        "В этом воркбуке мы познакомимся с библиотеками для обработки естественного языка и отработаем цикл FOR."
      ],
      "metadata": {
        "id": "T3ypiQzM_yon"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "**HuggingFace Datasets**\n",
        "\n",
        "HuggingFace - платформа для работы с моделями с архитектурой Transformer для обработки текстов и не только!\n",
        "\n",
        "Datasets дает доступ к коллекциям данных и позволяет создавать свои наборы для машинного обучения.\n",
        "\n",
        "`pip install` - команда shell для загрузки библиотеки\n",
        "\n",
        "`-q` - параметр \"quiet\", не выводит не экран подробную информацию о процессе загрузки библиотеки\n",
        "\n",
        "`!` - символ для объявления shell-команд"
      ],
      "metadata": {
        "id": "Zoavyuv3Bfax"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "!pip install -q datasets"
      ],
      "metadata": {
        "id": "IDkAQDPE4Hin"
      },
      "execution_count": 16,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "Загрузим датасет для анализа тональности текстов"
      ],
      "metadata": {
        "id": "dTikHZH4CaFo"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from datasets import load_dataset\n",
        "\n",
        "dataset = load_dataset(\"Sp1786/multiclass-sentiment-analysis-dataset\")"
      ],
      "metadata": {
        "id": "lx1PhjGlBUrK"
      },
      "execution_count": 17,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "Проведем анализ!\n",
        "\n",
        "Структура объекта dataset довольно сложная, но удобная для интерпретации.\n",
        "\n",
        "У нас есть 3 выборки: train, validation и test.\n",
        "\n",
        "В каждой выборке 4 столбца:\n",
        "\n",
        "- ID элемента датасета, он нужен, чтобы все значения были уникальными, как в реляционной БД\n",
        "- text, label, sentiment"
      ],
      "metadata": {
        "id": "g9CAEMZkCdX0"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "dataset"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "lE8JMz9fCfNn",
        "outputId": "44f462b9-3805-4a52-d0be-74e616e9b23d"
      },
      "execution_count": 18,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "DatasetDict({\n",
              "    train: Dataset({\n",
              "        features: ['id', 'text', 'label', 'sentiment'],\n",
              "        num_rows: 31232\n",
              "    })\n",
              "    validation: Dataset({\n",
              "        features: ['id', 'text', 'label', 'sentiment'],\n",
              "        num_rows: 5205\n",
              "    })\n",
              "    test: Dataset({\n",
              "        features: ['id', 'text', 'label', 'sentiment'],\n",
              "        num_rows: 5206\n",
              "    })\n",
              "})"
            ]
          },
          "metadata": {},
          "execution_count": 18
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Обычно для обработки таких данных используется библиотека pandas, но пока мы учимся, будем использовать нативный метод - цикл FOR.\n",
        "\n",
        "#**Задание 1**\n",
        "\n",
        "Для начала проверим, сбалансирован ли наш датасет."
      ],
      "metadata": {
        "id": "HsCrxJtaDAIn"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Какие метки есть в датасете?\n",
        "\n",
        "# Создаем множество set()\n",
        "# Множество - это список уникальных элементов\n",
        "# Этот тип данных можно использовать как фильтр для наших меток\n",
        "unique_labels = set()\n",
        "\n",
        "# Для каждого уникального элемента из обучающей выборки dataset['train']\n",
        "# Добавим элемент в множество unique_labels\n",
        "for i in dataset['train']:\n",
        "  unique_labels.add(i['sentiment'])\n",
        "\n",
        "unique_labels"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "IsKBV5wsCf7G",
        "outputId": "61afcbd1-d5da-4eed-fc0a-e80f8abbfc30"
      },
      "execution_count": 19,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "{'negative', 'neutral', 'positive'}"
            ]
          },
          "metadata": {},
          "execution_count": 19
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Теперь посчитаем, сколько представителей каждой метки в выборках train, validation и test\n",
        "neg_train = neg_val = neg_test = 0 # счетчики\n",
        "neut_train = neut_val = neut_test = 0\n",
        "pos_train = pos_val = pos_test = 0\n",
        "#обучающая выборка\n",
        "for item in dataset['train']:\n",
        "    sentiment = item['sentiment']\n",
        "    if sentiment == \"negative\":\n",
        "        neg_train += 1\n",
        "    elif sentiment == 'positive':\n",
        "        pos_train += 1\n",
        "    elif sentiment == 'neutral':\n",
        "        neut_train += 1\n",
        "\n",
        "#валидационная выборка\n",
        "for item in dataset.get('validation', []):\n",
        "    sentiment = item['sentiment']\n",
        "    if sentiment == \"negative\":\n",
        "        neg_val += 1\n",
        "    elif sentiment == 'positive':\n",
        "        pos_val += 1\n",
        "    elif sentiment == 'neutral':\n",
        "        neut_val += 1\n",
        "\n",
        "# тестовая выборка\n",
        "for item in dataset.get('test', []):\n",
        "    sentiment = item['sentiment']\n",
        "    if sentiment == \"negative\":\n",
        "        neg_test += 1\n",
        "    elif sentiment == 'positive':\n",
        "        pos_test += 1\n",
        "    elif sentiment == 'neutral':\n",
        "        neut_test += 1\n",
        "\n",
        "print('\\nДанные обучающей выборки:')\n",
        "print('Положительные метки:', pos_train)\n",
        "print('Отрицательные метки:', neg_train)\n",
        "print('Нейтральные метки:', neut_train)\n",
        "print('\\nДанные валидационной выборки:')\n",
        "print('Положительные метки:', pos_val)\n",
        "print('Отрицательные метки:', neg_val)\n",
        "print('Нейтральные метки:', neut_val)\n",
        "print('\\nДанные тестовой выборки:')\n",
        "print('Положительные метки:', pos_test)\n",
        "print('Отрицательные метки:', neg_test)\n",
        "print('Нейтральные метки:', neut_test)\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "KnPchjNYDOcu",
        "outputId": "a311bac1-7bee-43e7-8a37-5356b0997f10"
      },
      "execution_count": 20,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "Данные обучающей выборки:\n",
            "Положительные метки: 10478\n",
            "Отрицательные метки: 9105\n",
            "Нейтральные метки: 11649\n",
            "\n",
            "Данные валидационной выборки:\n",
            "Положительные метки: 1760\n",
            "Отрицательные метки: 1517\n",
            "Нейтральные метки: 1928\n",
            "\n",
            "Данные тестовой выборки:\n",
            "Положительные метки: 1730\n",
            "Отрицательные метки: 1546\n",
            "Нейтральные метки: 1930\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Используйте арифметические операции, чтобы посчитать процентное соотношение или долю каждого класса в каждой выборке\n",
        "perc_pos_train = pos_train / (pos_train + neg_train + neut_train)\n",
        "perc_neg_train = neg_train / (pos_train + neg_train + neut_train)\n",
        "perc_neut_train = neut_train / (pos_train + neg_train + neut_train)\n",
        "\n",
        "perc_pos_val = pos_val / (pos_val + neg_val + neut_val)\n",
        "perc_neg_val = neg_val / (pos_val + neg_val + neut_val)\n",
        "perc_neut_val = neut_val / (pos_val + neg_val + neut_val)\n",
        "\n",
        "perc_pos_test = pos_test / (pos_test + neg_test + neut_test)\n",
        "perc_neg_test = neg_test / (pos_test + neg_test + neut_test)\n",
        "perc_neut_test = neut_test / (pos_test + neg_test + neut_test)\n",
        "\n",
        "print(f'Положительные метки составляют {round(perc_pos_train * 100)} % обучающей выборки')\n",
        "print(f'Негативные метки составляют {round(perc_neg_train * 100)} % обучающей выборки')\n",
        "print(f'Нейтральные метки составляют {round(perc_neut_train * 100)} % обучающей выборки')\n",
        "\n",
        "print(f'\\nПоложительные метки составляют {round(perc_pos_val * 100)} % валидационной выборки')\n",
        "print(f'Негативные метки составляют {round(perc_neg_val * 100)} % валидационной выборки')\n",
        "print(f'Нейтральные метки составляют {round(perc_neut_val * 100)} % валидационной выборки')\n",
        "\n",
        "print(f'\\nПоложительные метки составляют {round(perc_pos_test * 100)} % тестовой выборки')\n",
        "print(f'Негативные метки составляют {round(perc_neg_test * 100)} % тестовой выборки')\n",
        "print(f'Нейтральные метки составляют {round(perc_neut_test * 100)} % тестовой выборки')\n",
        "\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "a8fO7lvcGR5T",
        "outputId": "f877e78a-4c17-456d-efa9-39e9379f776a"
      },
      "execution_count": 21,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Положительные метки составляют 34 % обучающей выборки\n",
            "Негативные метки составляют 29 % обучающей выборки\n",
            "Нейтральные метки составляют 37 % обучающей выборки\n",
            "\n",
            "Положительные метки составляют 34 % валидационной выборки\n",
            "Негативные метки составляют 29 % валидационной выборки\n",
            "Нейтральные метки составляют 37 % валидационной выборки\n",
            "\n",
            "Положительные метки составляют 33 % тестовой выборки\n",
            "Негативные метки составляют 30 % тестовой выборки\n",
            "Нейтральные метки составляют 37 % тестовой выборки\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Создадим визуализацию\n",
        "# Будем работать с библиотекой matplotlip\n",
        "# Numpy - библиотека для обработки массивов данных - обеспечивает работу matplotlib\n",
        "from matplotlib import pyplot as plt\n",
        "import numpy as np\n",
        "\n",
        "# Задаем список меток\n",
        "# Задаем значения для каждой метки, процентное соотношение (только для обучающей выборки)\n",
        "labels = ['Positive', 'Negative', 'Neutral']\n",
        "data = [perc_pos_train, perc_neg_train, perc_neut_train]\n",
        "\n",
        "# Код для отрисовки круговой диаграммы\n",
        "fig = plt.figure(figsize=(10, 7))\n",
        "plt.pie(data, labels=labels, autopct='%1.f%%') # autopict создает числовые подписи и использует целочисленное форматирование\n",
        "\n",
        "# Показать график\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 576
        },
        "id": "Ucjw-9whGqTw",
        "outputId": "ddc69773-c24e-4d3a-8793-c2b34c4c331a"
      },
      "execution_count": 22,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1000x700 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Сделаем то же самое для проверочной выборки\n",
        "labels = ['Positive', 'Negative', 'Neutral']\n",
        "data = [perc_pos_val, perc_neg_val, perc_neut_val]\n",
        "\n",
        "fig = plt.figure(figsize=(10, 7))\n",
        "plt.pie(data, labels=labels, autopct='%1.f%%')\n",
        "\n",
        "# Показать график\n",
        "plt.show()"
      ],
      "metadata": {
        "id": "q6cBY84aH2E0",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 576
        },
        "outputId": "c288f6ef-574c-49ce-d60e-c558bf375bc7"
      },
      "execution_count": 23,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1000x700 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Сделаем то же самое для тестовой выборки\n",
        "labels = ['Positive', 'Negative', 'Neutral']\n",
        "data = [perc_pos_test, perc_neg_test, perc_neut_test]\n",
        "\n",
        "fig = plt.figure(figsize=(10, 7))\n",
        "plt.pie(data, labels=labels, autopct='%1.f%%')\n",
        "\n",
        "# Показать график\n",
        "plt.show()"
      ],
      "metadata": {
        "id": "P6kstLAHH7kJ",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 576
        },
        "outputId": "0b97ab62-5bf8-4267-d275-58a304bf4f10"
      },
      "execution_count": 24,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1000x700 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "#**Задание 2**\n",
        "\n",
        "Займемся обработкой естественного языка.\n",
        "\n",
        "Познакомимся с библиотекой SpaCy для обработки текстов."
      ],
      "metadata": {
        "id": "FXJuNfwVGBBc"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Загрузка spaCy\n",
        "import spacy\n",
        "\n",
        "# Загрузка библиотеки для визуализации spaCy\n",
        "from spacy import displacy\n",
        "\n",
        "# Инициализация инструментов SpaCy для английского (для каждого языка загружается свой вариант)\n",
        "nlp = spacy.load('en_core_web_sm')\n",
        "\n",
        "# Выведем доступные инструменты\n",
        "print(nlp.pipe_names)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "TZDla6PuaF4C",
        "outputId": "f35cad3c-d76b-4ba2-8d79-20928f52bb6f"
      },
      "execution_count": 25,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "['tok2vec', 'tagger', 'parser', 'attribute_ruler', 'lemmatizer', 'ner']\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Токенизация SpaCy производится с помощью цикла for"
      ],
      "metadata": {
        "id": "dee2sTbia0vF"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Образец текста из нашего датасета\n",
        "sample = dataset['train'][0]['text']\n",
        "sample"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 35
        },
        "id": "2mDcYdu4bCOQ",
        "outputId": "eaa1c25a-d5bd-4a13-8e4c-50cfe1114742"
      },
      "execution_count": 26,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "'Cooking microwave pizzas, yummy'"
            ],
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "string"
            }
          },
          "metadata": {},
          "execution_count": 26
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Для работы со SpaCy, _нужно_ привести все к нужному виду\n",
        "# Для этого загружаем в nlp наш текст\n",
        "doc = nlp(sample)\n",
        "doc # Визуально ничего не изменилось, но теперь это объект класса SpaCy, который можно обработать инструментами NLP"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "RnqH0yzEbHKC",
        "outputId": "570e98ee-e469-498e-cf93-9c9bc9d6392b"
      },
      "execution_count": 27,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Cooking microwave pizzas, yummy"
            ]
          },
          "metadata": {},
          "execution_count": 27
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Токенизация\n",
        "for token in doc:\n",
        "    print(token.text)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Gx0LPZAUZ0A0",
        "outputId": "cf8cdcbd-8cf4-47a3-ebfb-75b0b7b0f85b"
      },
      "execution_count": 28,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Cooking\n",
            "microwave\n",
            "pizzas\n",
            ",\n",
            "yummy\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Частересная разметка\n",
        "for token in doc:\n",
        "    print(token.pos_)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "pQc83BZGbfk6",
        "outputId": "3388c52b-1816-4022-f989-5fecc51b9379"
      },
      "execution_count": 29,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "NOUN\n",
            "NOUN\n",
            "NOUN\n",
            "PUNCT\n",
            "ADJ\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Лемматизация\n",
        "for token in doc:\n",
        "    print(token.lemma_)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ZWgH4deYbpjJ",
        "outputId": "6e3ef621-ae6c-43fb-d04b-ed0f772b660c"
      },
      "execution_count": 30,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "cooking\n",
            "microwave\n",
            "pizza\n",
            ",\n",
            "yummy\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Синтаксическая роль (составляющие)\n",
        "for token in doc:\n",
        "    print(token.tag_)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "DUseEIwMbzrW",
        "outputId": "76a3bb3f-b5fb-413f-9c2f-47954d911d21"
      },
      "execution_count": 31,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "NN\n",
            "NN\n",
            "NNS\n",
            ",\n",
            "JJ\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Синтаксическая роль (зависимости)\n",
        "for token in doc:\n",
        "\n",
        "    print(token.dep_)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "PoiI91_2b361",
        "outputId": "b5d7580c-6902-47b8-d16c-e8457a8651cd"
      },
      "execution_count": 32,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "compound\n",
            "compound\n",
            "ROOT\n",
            "punct\n",
            "amod\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Морфологическая разметка\n",
        "for token in doc:\n",
        "    print(token.morph)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "WwAJ1_U6cTfX",
        "outputId": "c2c6d095-760a-40f5-907c-33f1c7cc0a50"
      },
      "execution_count": 33,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Number=Sing\n",
            "Number=Sing\n",
            "Number=Plur\n",
            "PunctType=Comm\n",
            "Degree=Pos\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Задание!\n",
        "\n",
        "1. Сохраним в отдельную переменную список текстов из тестовой выборки"
      ],
      "metadata": {
        "id": "D6E9WuBOdFMc"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "### ваш код здесь ###\n",
        "sample = dataset['test'][:5]['text']\n",
        "print(sample)\n",
        "# Выведите первые 5 текстов"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "J6O0t2zhcb_G",
        "outputId": "e0c52574-557d-410d-bc85-0ff3a781b98e"
      },
      "execution_count": 34,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "['getting cds ready for tour', ' MC, happy mother`s day to your mom ;).. love yah', 'A year from now is graduation....i am pretty sure i`m not ready for it!?!?!?', ' because you had chips and sale w/o me', 'Great for organising my work life balance']\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "2. Создадим список токенов для каждого текста с помощью SpaCy"
      ],
      "metadata": {
        "id": "eRgN1Qr4dnZT"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "tokens = [[token.text for token in nlp(text)] for text in sample] # Создаем список частей речи для каждого токена каждого текста\n",
        "tokens\n",
        "# Обработка может занять много времени, достаточно обработать первые 100 текстов"
      ],
      "metadata": {
        "id": "SXsyPzricnwo",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "dd30453f-0c56-411f-b2a8-4a6e397160c5"
      },
      "execution_count": 35,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[['getting', 'cds', 'ready', 'for', 'tour'],\n",
              " [' ',\n",
              "  'MC',\n",
              "  ',',\n",
              "  'happy',\n",
              "  'mother`s',\n",
              "  'day',\n",
              "  'to',\n",
              "  'your',\n",
              "  'mom',\n",
              "  ';)',\n",
              "  '..',\n",
              "  'love',\n",
              "  'yah'],\n",
              " ['A',\n",
              "  'year',\n",
              "  'from',\n",
              "  'now',\n",
              "  'is',\n",
              "  'graduation',\n",
              "  '....',\n",
              "  'i',\n",
              "  'am',\n",
              "  'pretty',\n",
              "  'sure',\n",
              "  'i`m',\n",
              "  'not',\n",
              "  'ready',\n",
              "  'for',\n",
              "  'it',\n",
              "  '!',\n",
              "  '?',\n",
              "  '!',\n",
              "  '?',\n",
              "  '!',\n",
              "  '?'],\n",
              " [' ', 'because', 'you', 'had', 'chips', 'and', 'sale', 'w/o', 'me'],\n",
              " ['Great', 'for', 'organising', 'my', 'work', 'life', 'balance']]"
            ]
          },
          "metadata": {},
          "execution_count": 35
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "3. Частеречная разметка: создаем список частей речи для каждого токена"
      ],
      "metadata": {
        "id": "OXzB_0AifeeT"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "pos = [[token.pos_ for token in nlp(text)] for text in sample]\n",
        "pos"
      ],
      "metadata": {
        "id": "RZI3ZOOGePY2",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "2a133830-b651-4281-d19b-63d9e3fa37a5"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[['VERB', 'NOUN', 'ADJ', 'ADP', 'NOUN'],\n",
              " ['SPACE',\n",
              "  'PROPN',\n",
              "  'PUNCT',\n",
              "  'ADJ',\n",
              "  'PROPN',\n",
              "  'NOUN',\n",
              "  'ADP',\n",
              "  'PRON',\n",
              "  'NOUN',\n",
              "  'PUNCT',\n",
              "  'PUNCT',\n",
              "  'NOUN',\n",
              "  'NOUN'],\n",
              " ['DET',\n",
              "  'NOUN',\n",
              "  'ADP',\n",
              "  'ADV',\n",
              "  'AUX',\n",
              "  'NOUN',\n",
              "  'PUNCT',\n",
              "  'PRON',\n",
              "  'AUX',\n",
              "  'ADV',\n",
              "  'ADJ',\n",
              "  'PROPN',\n",
              "  'PART',\n",
              "  'ADJ',\n",
              "  'ADP',\n",
              "  'PRON',\n",
              "  'PUNCT',\n",
              "  'PUNCT',\n",
              "  'PUNCT',\n",
              "  'PUNCT',\n",
              "  'PUNCT',\n",
              "  'PUNCT'],\n",
              " ['SPACE', 'SCONJ', 'PRON', 'VERB', 'NOUN', 'CCONJ', 'NOUN', 'ADP', 'PRON'],\n",
              " ['ADJ', 'ADP', 'VERB', 'PRON', 'NOUN', 'NOUN', 'NOUN'],\n",
              " ['SPACE',\n",
              "  'PRON',\n",
              "  'PRON',\n",
              "  'VERB',\n",
              "  'ADV',\n",
              "  'NOUN',\n",
              "  'SPACE',\n",
              "  'PUNCT',\n",
              "  'PROPN',\n",
              "  'SPACE',\n",
              "  'PRON',\n",
              "  'AUX',\n",
              "  'VERB',\n",
              "  'PUNCT'],\n",
              " ['AUX',\n",
              "  'PROPN',\n",
              "  'PROPN',\n",
              "  'PROPN',\n",
              "  'NOUN',\n",
              "  'PUNCT',\n",
              "  'CCONJ',\n",
              "  'AUX',\n",
              "  'VERB',\n",
              "  'PRON',\n",
              "  'NOUN',\n",
              "  'PUNCT',\n",
              "  'INTJ',\n",
              "  'PROPN',\n",
              "  'VERB',\n",
              "  'ADP',\n",
              "  'ADJ',\n",
              "  'NOUN',\n",
              "  'NOUN'],\n",
              " ['SPACE',\n",
              "  'VERB',\n",
              "  'DET',\n",
              "  'NOUN',\n",
              "  'PRON',\n",
              "  'VERB',\n",
              "  'X',\n",
              "  'VERB',\n",
              "  'PRON',\n",
              "  'PUNCT',\n",
              "  'PUNCT',\n",
              "  'X',\n",
              "  'PUNCT',\n",
              "  'NUM',\n",
              "  'SPACE',\n",
              "  'VERB',\n",
              "  'PRON',\n",
              "  'VERB',\n",
              "  'NUM',\n",
              "  'NOUN',\n",
              "  'CCONJ',\n",
              "  'VERB',\n",
              "  'PRON',\n",
              "  'VERB',\n",
              "  'PUNCT'],\n",
              " ['SPACE',\n",
              "  'PRON',\n",
              "  'PUNCT',\n",
              "  'ADV',\n",
              "  'PRON',\n",
              "  'AUX',\n",
              "  'PRON',\n",
              "  'ADJ',\n",
              "  'NOUN',\n",
              "  'PUNCT',\n",
              "  'SPACE',\n",
              "  'PART',\n",
              "  'PRON',\n",
              "  'AUX',\n",
              "  'ADV',\n",
              "  'ADJ',\n",
              "  'ADP',\n",
              "  'SPACE',\n",
              "  'PUNCT',\n",
              "  'ADJ',\n",
              "  'NOUN',\n",
              "  'NOUN',\n",
              "  'PUNCT'],\n",
              " ['NOUN',\n",
              "  'PRON',\n",
              "  'AUX',\n",
              "  'ADV',\n",
              "  'ADJ',\n",
              "  'VERB',\n",
              "  'PRON',\n",
              "  'VERB',\n",
              "  'ADP',\n",
              "  'NOUN',\n",
              "  'ADV',\n",
              "  'PUNCT',\n",
              "  'CCONJ',\n",
              "  'ADV',\n",
              "  'PROPN',\n",
              "  'PROPN',\n",
              "  'PUNCT',\n",
              "  'PUNCT'],\n",
              " ['SPACE',\n",
              "  'CCONJ',\n",
              "  'CCONJ',\n",
              "  'PRON',\n",
              "  'AUX',\n",
              "  'ADJ',\n",
              "  'SCONJ',\n",
              "  'PRON',\n",
              "  'VERB',\n",
              "  'PRON'],\n",
              " ['SPACE',\n",
              "  'PRON',\n",
              "  'VERB',\n",
              "  'DET',\n",
              "  'DET',\n",
              "  'ADJ',\n",
              "  'PROPN',\n",
              "  'ADP',\n",
              "  'DET',\n",
              "  'NOUN',\n",
              "  'PUNCT',\n",
              "  'PRON',\n",
              "  'VERB',\n",
              "  'INTJ',\n",
              "  'ADP',\n",
              "  'PRON',\n",
              "  'VERB',\n",
              "  'PRON',\n",
              "  'AUX',\n",
              "  'PRON',\n",
              "  'PUNCT',\n",
              "  'PRON',\n",
              "  'NOUN'],\n",
              " ['PRON',\n",
              "  'VERB',\n",
              "  'DET',\n",
              "  'NOUN',\n",
              "  'PUNCT',\n",
              "  'VERB',\n",
              "  'ADV',\n",
              "  'ADJ',\n",
              "  'PRON',\n",
              "  'VERB',\n",
              "  'PRON',\n",
              "  'PUNCT',\n",
              "  'ADV',\n",
              "  'PRON',\n",
              "  'AUX',\n",
              "  'VERB',\n",
              "  'PRON',\n",
              "  'ADP',\n",
              "  'CCONJ',\n",
              "  'VERB',\n",
              "  'PRON',\n",
              "  'AUX',\n",
              "  'PART',\n",
              "  'VERB',\n",
              "  'PUNCT'],\n",
              " ['SPACE',\n",
              "  'PROPN',\n",
              "  'PUNCT',\n",
              "  'PUNCT',\n",
              "  'INTJ',\n",
              "  'PUNCT',\n",
              "  'INTJ',\n",
              "  'PRON',\n",
              "  'DET',\n",
              "  'SPACE',\n",
              "  'NOUN'],\n",
              " ['SPACE',\n",
              "  'PROPN',\n",
              "  'ADV',\n",
              "  'VERB',\n",
              "  'PART',\n",
              "  'VERB',\n",
              "  'NOUN',\n",
              "  'PART',\n",
              "  'ADV',\n",
              "  'VERB',\n",
              "  'DET',\n",
              "  'NOUN',\n",
              "  'ADP',\n",
              "  'NOUN',\n",
              "  'DET',\n",
              "  'NOUN'],\n",
              " ['NOUN',\n",
              "  'AUX',\n",
              "  'ADV',\n",
              "  'ADJ',\n",
              "  'PUNCT',\n",
              "  'PRON',\n",
              "  'ADV',\n",
              "  'VERB',\n",
              "  'DET',\n",
              "  'NOUN',\n",
              "  'SCONJ',\n",
              "  'VERB',\n",
              "  'DET',\n",
              "  'NOUN',\n",
              "  'VERB',\n",
              "  'PRON',\n",
              "  'AUX',\n",
              "  'ADV',\n",
              "  'ADJ',\n",
              "  'PUNCT',\n",
              "  'ADV',\n",
              "  'PUNCT',\n",
              "  'DET',\n",
              "  'NOUN',\n",
              "  'AUX',\n",
              "  'VERB',\n",
              "  'ADP',\n",
              "  'ADV',\n",
              "  'PUNCT',\n",
              "  'VERB',\n",
              "  'PRON',\n",
              "  'AUX',\n",
              "  'VERB',\n",
              "  'PUNCT'],\n",
              " ['ADP', 'PART', 'VERB', 'ADP', 'DET', 'NOUN'],\n",
              " ['INTJ',\n",
              "  'PRON',\n",
              "  'VERB',\n",
              "  'PUNCT',\n",
              "  'ADV',\n",
              "  'ADJ',\n",
              "  'ADP',\n",
              "  'NOUN',\n",
              "  'NOUN',\n",
              "  'ADP',\n",
              "  'PRON',\n",
              "  'ADJ',\n",
              "  'NOUN',\n",
              "  'PUNCT',\n",
              "  'PROPN',\n",
              "  'PUNCT'],\n",
              " ['SPACE',\n",
              "  'PRON',\n",
              "  'ADV',\n",
              "  'AUX',\n",
              "  'PUNCT',\n",
              "  'SCONJ',\n",
              "  'ADP',\n",
              "  'DET',\n",
              "  'NOUN',\n",
              "  'SPACE',\n",
              "  'SCONJ',\n",
              "  'PRON',\n",
              "  'ADV',\n",
              "  'ADV',\n",
              "  'VERB',\n",
              "  'ADJ',\n",
              "  'NOUN',\n",
              "  'PUNCT',\n",
              "  'CCONJ',\n",
              "  'VERB',\n",
              "  'ADJ',\n",
              "  'ADP',\n",
              "  'ADJ',\n",
              "  'NOUN',\n",
              "  'VERB',\n",
              "  'NOUN'],\n",
              " ['ADP',\n",
              "  'DET',\n",
              "  'NOUN',\n",
              "  'CCONJ',\n",
              "  'PRON',\n",
              "  'VERB',\n",
              "  'ADP',\n",
              "  'ADV',\n",
              "  'ADV',\n",
              "  'PUNCT',\n",
              "  'ADV',\n",
              "  'ADP',\n",
              "  'VERB',\n",
              "  'DET',\n",
              "  'NOUN',\n",
              "  'DET',\n",
              "  'NOUN',\n",
              "  'VERB',\n",
              "  'CCONJ',\n",
              "  'VERB',\n",
              "  'PRON',\n",
              "  'ADV',\n",
              "  'PUNCT',\n",
              "  'AUX',\n",
              "  'PART',\n",
              "  'VERB',\n",
              "  'PUNCT',\n",
              "  'PUNCT'],\n",
              " ['NOUN',\n",
              "  'ADP',\n",
              "  'NOUN',\n",
              "  'PART',\n",
              "  'VERB',\n",
              "  'NOUN',\n",
              "  'PUNCT',\n",
              "  'ADV',\n",
              "  'PUNCT',\n",
              "  'VERB',\n",
              "  'ADJ',\n",
              "  'NOUN',\n",
              "  'ADP',\n",
              "  'PROPN',\n",
              "  'PROPN',\n",
              "  'PROPN',\n",
              "  'NOUN',\n",
              "  'PUNCT',\n",
              "  'X',\n",
              "  'PUNCT',\n",
              "  'CCONJ',\n",
              "  'NOUN'],\n",
              " ['SPACE',\n",
              "  'INTJ',\n",
              "  'PRON',\n",
              "  'VERB',\n",
              "  'PRON',\n",
              "  'PUNCT',\n",
              "  'SPACE',\n",
              "  'NOUN',\n",
              "  'SPACE',\n",
              "  'VERB',\n",
              "  'PRON',\n",
              "  'VERB',\n",
              "  'ADV',\n",
              "  'PUNCT',\n",
              "  'PRON',\n",
              "  'VERB',\n",
              "  'PRON',\n",
              "  'AUX',\n",
              "  'PART',\n",
              "  'VERB',\n",
              "  'ADV',\n",
              "  'PUNCT'],\n",
              " ['ADV', 'PUNCT', 'PRON', 'AUX', 'PART', 'VERB', 'ADJ'],\n",
              " ['NUM',\n",
              "  'PROPN',\n",
              "  'ADP',\n",
              "  'PRON',\n",
              "  'PRON',\n",
              "  'VERB',\n",
              "  'ADV',\n",
              "  'PUNCT',\n",
              "  'SPACE',\n",
              "  'ADJ',\n",
              "  'PUNCT',\n",
              "  'NOUN',\n",
              "  'ADP',\n",
              "  'PRON',\n",
              "  'PRON',\n",
              "  'PROPN',\n",
              "  'PUNCT',\n",
              "  'SCONJ',\n",
              "  'PRON',\n",
              "  'VERB',\n",
              "  'ADP',\n",
              "  'AUX',\n",
              "  'ADJ',\n",
              "  'PART',\n",
              "  'VERB',\n",
              "  'NOUN',\n",
              "  'PRON',\n",
              "  'VERB',\n",
              "  'PART',\n",
              "  'SCONJ',\n",
              "  'VERB',\n",
              "  'ADP',\n",
              "  'NUM',\n",
              "  'NUM'],\n",
              " ['SPACE',\n",
              "  'PRON',\n",
              "  'ADV',\n",
              "  'VERB',\n",
              "  'PRON',\n",
              "  'CCONJ',\n",
              "  'ADV',\n",
              "  'DET',\n",
              "  'NOUN',\n",
              "  'AUX',\n",
              "  'PRON',\n",
              "  'VERB'],\n",
              " ['DET',\n",
              "  'NOUN',\n",
              "  'NOUN',\n",
              "  'ADJ',\n",
              "  'ADP',\n",
              "  'NOUN',\n",
              "  'PUNCT',\n",
              "  'DET',\n",
              "  'NOUN',\n",
              "  'ADJ',\n",
              "  'PUNCT',\n",
              "  'DET',\n",
              "  'NOUN',\n",
              "  'ADP',\n",
              "  'VERB',\n",
              "  'NOUN',\n",
              "  'NOUN',\n",
              "  'PUNCT',\n",
              "  'ADV',\n",
              "  'VERB',\n",
              "  'DET',\n",
              "  'NOUN',\n",
              "  'PART',\n",
              "  'VERB',\n",
              "  'ADJ',\n",
              "  'NOUN',\n",
              "  'PUNCT'],\n",
              " ['PROPN',\n",
              "  'AUX',\n",
              "  'VERB',\n",
              "  'DET',\n",
              "  'ADJ',\n",
              "  'ADJ',\n",
              "  'NOUN',\n",
              "  'PRON',\n",
              "  'VERB',\n",
              "  'VERB',\n",
              "  'NOUN'],\n",
              " ['VERB',\n",
              "  'NUM',\n",
              "  'ADP',\n",
              "  'NOUN',\n",
              "  'NOUN',\n",
              "  'PUNCT',\n",
              "  'NUM',\n",
              "  'NOUN',\n",
              "  'NOUN',\n",
              "  'CCONJ',\n",
              "  'SCONJ',\n",
              "  'PRON',\n",
              "  'AUX',\n",
              "  'PART',\n",
              "  'VERB',\n",
              "  'ADP',\n",
              "  'PRON',\n",
              "  'PUNCT',\n",
              "  'PRON',\n",
              "  'AUX',\n",
              "  'VERB',\n",
              "  'NUM',\n",
              "  'ADJ',\n",
              "  'ADP',\n",
              "  'NOUN'],\n",
              " ['SPACE',\n",
              "  'PRON',\n",
              "  'AUX',\n",
              "  'DET',\n",
              "  'ADJ',\n",
              "  'NOUN',\n",
              "  'ADP',\n",
              "  'DET',\n",
              "  'NOUN',\n",
              "  'SPACE',\n",
              "  'SCONJ',\n",
              "  'PRON',\n",
              "  'VERB',\n",
              "  'ADP',\n",
              "  'PROPN',\n",
              "  'NOUN',\n",
              "  'PUNCT'],\n",
              " ['SPACE',\n",
              "  'INTJ',\n",
              "  'NOUN',\n",
              "  'SPACE',\n",
              "  'PROPN',\n",
              "  'NOUN',\n",
              "  'PUNCT',\n",
              "  'PUNCT',\n",
              "  'PUNCT',\n",
              "  'NOUN',\n",
              "  'NOUN',\n",
              "  'SCONJ',\n",
              "  'PRON',\n",
              "  'AUX',\n",
              "  'VERB',\n",
              "  'ADP',\n",
              "  'DET',\n",
              "  'NOUN',\n",
              "  'VERB',\n",
              "  'PRON',\n",
              "  'ADV',\n",
              "  'SPACE',\n",
              "  'NOUN'],\n",
              " ['INTJ',\n",
              "  'INTJ',\n",
              "  'NOUN',\n",
              "  'ADP',\n",
              "  'NOUN',\n",
              "  'ADV',\n",
              "  'PUNCT',\n",
              "  'PROPN',\n",
              "  'PUNCT',\n",
              "  'SPACE',\n",
              "  'PUNCT',\n",
              "  'PUNCT',\n",
              "  'NOUN',\n",
              "  'ADP',\n",
              "  'NOUN',\n",
              "  'PUNCT',\n",
              "  'PUNCT'],\n",
              " ['PROPN',\n",
              "  'NOUN',\n",
              "  'PRON',\n",
              "  'VERB',\n",
              "  'PRON',\n",
              "  'PROPN',\n",
              "  'VERB',\n",
              "  'ADJ',\n",
              "  'SCONJ',\n",
              "  'INTJ',\n",
              "  'DET',\n",
              "  'NOUN',\n",
              "  'SPACE',\n",
              "  'NOUN',\n",
              "  'PUNCT'],\n",
              " ['SPACE', 'PRON', 'AUX', 'ADV', 'DET', 'ADJ', 'ADJ', 'NOUN'],\n",
              " ['NOUN',\n",
              "  'PUNCT',\n",
              "  'PRON',\n",
              "  'VERB',\n",
              "  'PROPN',\n",
              "  'VERB',\n",
              "  'ADP',\n",
              "  'ADV',\n",
              "  'PUNCT',\n",
              "  'ADJ',\n",
              "  'NOUN'],\n",
              " ['NOUN',\n",
              "  'AUX',\n",
              "  'ADJ',\n",
              "  'PUNCT',\n",
              "  'ADJ',\n",
              "  'PUNCT',\n",
              "  'CCONJ',\n",
              "  'AUX',\n",
              "  'ADV',\n",
              "  'VERB',\n",
              "  'PRON',\n",
              "  'ADP',\n",
              "  'DET',\n",
              "  'NOUN',\n",
              "  'ADP',\n",
              "  'VERB',\n",
              "  'PRON',\n",
              "  'NOUN',\n",
              "  'ADP',\n",
              "  'VERB',\n",
              "  'NOUN',\n",
              "  'PUNCT',\n",
              "  'DET',\n",
              "  'ADJ',\n",
              "  'NOUN',\n",
              "  'AUX',\n",
              "  'DET',\n",
              "  'NOUN',\n",
              "  'PART',\n",
              "  'VERB',\n",
              "  'ADJ',\n",
              "  'NOUN',\n",
              "  'PUNCT',\n",
              "  'PROPN',\n",
              "  'PUNCT',\n",
              "  'VERB',\n",
              "  'PART',\n",
              "  'VERB',\n",
              "  'NOUN',\n",
              "  'PUNCT'],\n",
              " ['NOUN',\n",
              "  'AUX',\n",
              "  'ADJ',\n",
              "  'SCONJ',\n",
              "  'PRON',\n",
              "  'AUX',\n",
              "  'PART',\n",
              "  'ADP',\n",
              "  'PROPN',\n",
              "  'SPACE',\n",
              "  'PRON',\n",
              "  'AUX',\n",
              "  'ADJ',\n",
              "  'VERB',\n",
              "  'ADV',\n",
              "  'ADP',\n",
              "  'PROPN',\n",
              "  'PROPN'],\n",
              " ['NOUN',\n",
              "  'PRON',\n",
              "  'PUNCT',\n",
              "  'VERB',\n",
              "  'PUNCT',\n",
              "  'ADP',\n",
              "  'VERB',\n",
              "  'PRON',\n",
              "  'ADJ',\n",
              "  'NOUN',\n",
              "  'ADP',\n",
              "  'PROPN',\n",
              "  'AUX',\n",
              "  'PART',\n",
              "  'ADJ',\n",
              "  'PUNCT',\n",
              "  'PROPN',\n",
              "  'NOUN',\n",
              "  'PUNCT',\n",
              "  'NUM',\n",
              "  'NOUN',\n",
              "  'NOUN',\n",
              "  'ADP',\n",
              "  'DET',\n",
              "  'NUM',\n",
              "  'NOUN',\n",
              "  'NOUN',\n",
              "  'PUNCT',\n",
              "  'DET',\n",
              "  'NOUN',\n",
              "  'PUNCT'],\n",
              " ['SPACE', 'NOUN', 'PUNCT', 'SCONJ', 'PUNCT'],\n",
              " ['VERB',\n",
              "  'ADJ',\n",
              "  'CCONJ',\n",
              "  'DET',\n",
              "  'NOUN',\n",
              "  'VERB',\n",
              "  'DET',\n",
              "  'NOUN',\n",
              "  'PRON',\n",
              "  'ADJ',\n",
              "  'NOUN',\n",
              "  'AUX',\n",
              "  'PART',\n",
              "  'VERB',\n",
              "  'PUNCT',\n",
              "  'PRON',\n",
              "  'AUX',\n",
              "  'VERB',\n",
              "  'NOUN',\n",
              "  'PRON',\n",
              "  'PRON',\n",
              "  'VERB',\n",
              "  'PUNCT',\n",
              "  'PROPN',\n",
              "  'PROPN',\n",
              "  'PUNCT'],\n",
              " ['SPACE',\n",
              "  'PROPN',\n",
              "  'PRON',\n",
              "  'VERB',\n",
              "  'VERB',\n",
              "  'DET',\n",
              "  'NOUN',\n",
              "  'SCONJ',\n",
              "  'ADV',\n",
              "  'PUNCT',\n",
              "  'SCONJ',\n",
              "  'PRON',\n",
              "  'VERB',\n",
              "  'ADV',\n",
              "  'VERB',\n",
              "  'PRON',\n",
              "  'PRON',\n",
              "  'AUX',\n",
              "  'PART',\n",
              "  'VERB',\n",
              "  'DET',\n",
              "  'NOUN',\n",
              "  'NOUN',\n",
              "  'VERB',\n",
              "  'ADP',\n",
              "  'PRON',\n",
              "  'NOUN'],\n",
              " ['ADJ', 'NOUN', 'ADP', 'DET', 'NOUN', 'SPACE', 'PUNCT'],\n",
              " ['SPACE',\n",
              "  'PROPN',\n",
              "  'VERB',\n",
              "  'ADP',\n",
              "  'PRON',\n",
              "  'NOUN',\n",
              "  'PUNCT',\n",
              "  'PRON',\n",
              "  'VERB',\n",
              "  'NOUN',\n",
              "  'PUNCT',\n",
              "  'VERB',\n",
              "  'PART',\n",
              "  'VERB',\n",
              "  'PUNCT',\n",
              "  'CCONJ',\n",
              "  'ADV',\n",
              "  'VERB',\n",
              "  'ADJ'],\n",
              " ['VERB', 'ADV', 'PUNCT', 'DET', 'NOUN', 'NOUN'],\n",
              " ['SPACE',\n",
              "  'NOUN',\n",
              "  'PROPN',\n",
              "  'NOUN',\n",
              "  'NOUN',\n",
              "  'PRON',\n",
              "  'VERB',\n",
              "  'PRON',\n",
              "  'PROPN',\n",
              "  'NOUN',\n",
              "  'ADJ',\n",
              "  'NOUN',\n",
              "  'ADV',\n",
              "  'PROPN',\n",
              "  'PROPN',\n",
              "  'PROPN',\n",
              "  'CCONJ',\n",
              "  'DET',\n",
              "  'PUNCT',\n",
              "  'NOUN',\n",
              "  'PUNCT',\n",
              "  'NOUN'],\n",
              " ['SPACE',\n",
              "  'PROPN',\n",
              "  'VERB',\n",
              "  'DET',\n",
              "  'NOUN',\n",
              "  'ADP',\n",
              "  'ADV',\n",
              "  'PUNCT',\n",
              "  'CCONJ',\n",
              "  'PRON',\n",
              "  'VERB',\n",
              "  'PROPN',\n",
              "  'SPACE',\n",
              "  'NOUN',\n",
              "  'PART',\n",
              "  'VERB',\n",
              "  'ADP',\n",
              "  'ADV',\n",
              "  'PRON',\n",
              "  'NOUN',\n",
              "  'NOUN',\n",
              "  'AUX',\n",
              "  'ADJ'],\n",
              " ['ADJ',\n",
              "  'CCONJ',\n",
              "  'PROPN',\n",
              "  'ADJ',\n",
              "  'PROPN',\n",
              "  'NOUN',\n",
              "  'PUNCT',\n",
              "  'SPACE',\n",
              "  'CCONJ',\n",
              "  'PROPN',\n",
              "  'AUX',\n",
              "  'DET',\n",
              "  'ADJ',\n",
              "  'NOUN',\n",
              "  'SPACE',\n",
              "  'ADJ',\n",
              "  'ADP',\n",
              "  'DET',\n",
              "  'NOUN',\n",
              "  'PUNCT',\n",
              "  'SPACE',\n",
              "  'NOUN'],\n",
              " ['ADJ',\n",
              "  'PROPN',\n",
              "  'NOUN',\n",
              "  'CCONJ',\n",
              "  'ADP',\n",
              "  'DET',\n",
              "  'PROPN',\n",
              "  'PROPN',\n",
              "  'CCONJ',\n",
              "  'VERB',\n",
              "  'PROPN',\n",
              "  'ADV',\n",
              "  'PUNCT'],\n",
              " ['PUNCT',\n",
              "  'PROPN',\n",
              "  'PROPN',\n",
              "  'PROPN',\n",
              "  'PUNCT',\n",
              "  'PROPN',\n",
              "  'PUNCT',\n",
              "  'PUNCT',\n",
              "  'PUNCT',\n",
              "  'PRON',\n",
              "  'VERB',\n",
              "  'PROPN',\n",
              "  'SPACE',\n",
              "  'NOUN'],\n",
              " ['ADP',\n",
              "  'PRON',\n",
              "  'NOUN',\n",
              "  'PROPN',\n",
              "  'SPACE',\n",
              "  'PRON',\n",
              "  'VERB',\n",
              "  'ADJ',\n",
              "  'NOUN',\n",
              "  'NOUN'],\n",
              " ['ADV'],\n",
              " ['PRON', 'VERB', 'ADJ'],\n",
              " ['PUNCT',\n",
              "  'X',\n",
              "  'PRON',\n",
              "  'SCONJ',\n",
              "  'PRON',\n",
              "  'VERB',\n",
              "  'PRON',\n",
              "  'AUX',\n",
              "  'DET',\n",
              "  'PROPN',\n",
              "  'PROPN',\n",
              "  'NOUN',\n",
              "  'CCONJ',\n",
              "  'PRON',\n",
              "  'VERB',\n",
              "  'DET',\n",
              "  'ADV',\n",
              "  'ADJ',\n",
              "  'ADP',\n",
              "  'DET',\n",
              "  'ADJ',\n",
              "  'NOUN',\n",
              "  'PUNCT'],\n",
              " ['ADJ', 'NOUN', 'VERB', 'NOUN', 'NOUN', 'PUNCT', 'SPACE', 'PROPN'],\n",
              " ['AUX',\n",
              "  'PART',\n",
              "  'VERB',\n",
              "  'PRON',\n",
              "  'VERB',\n",
              "  'DET',\n",
              "  'ADJ',\n",
              "  'NOUN',\n",
              "  'ADP',\n",
              "  'VERB',\n",
              "  'PRON',\n",
              "  'ADP',\n",
              "  'PRON',\n",
              "  'PUNCT'],\n",
              " ['VERB',\n",
              "  'DET',\n",
              "  'ADJ',\n",
              "  'NOUN',\n",
              "  'ADP',\n",
              "  'DET',\n",
              "  'NOUN',\n",
              "  'PUNCT',\n",
              "  'ADJ',\n",
              "  'PART',\n",
              "  'AUX',\n",
              "  'ADV',\n",
              "  'ADP',\n",
              "  'NOUN',\n",
              "  'ADP',\n",
              "  'DET',\n",
              "  'NOUN'],\n",
              " ['INTJ',\n",
              "  'ADV',\n",
              "  'PRON',\n",
              "  'DET',\n",
              "  'NOUN',\n",
              "  'PRON',\n",
              "  'ADV',\n",
              "  'VERB',\n",
              "  'ADP',\n",
              "  'VERB',\n",
              "  'CCONJ',\n",
              "  'VERB',\n",
              "  'DET',\n",
              "  'ADJ'],\n",
              " ['PRON',\n",
              "  'NOUN',\n",
              "  'AUX',\n",
              "  'PART',\n",
              "  'VERB',\n",
              "  'ADP',\n",
              "  'DET',\n",
              "  'NOUN',\n",
              "  'PUNCT',\n",
              "  'SCONJ',\n",
              "  'AUX',\n",
              "  'PRON',\n",
              "  'VERB',\n",
              "  'PRON',\n",
              "  'PART',\n",
              "  'VERB',\n",
              "  'PUNCT',\n",
              "  'PRON',\n",
              "  'AUX',\n",
              "  'ADJ',\n",
              "  'ADV',\n",
              "  'PUNCT',\n",
              "  'SCONJ',\n",
              "  'AUX',\n",
              "  'PRON',\n",
              "  'VERB',\n",
              "  'DET',\n",
              "  'PRON',\n",
              "  'NOUN',\n",
              "  'ADP',\n",
              "  'DET',\n",
              "  'PROPN',\n",
              "  'NOUN',\n",
              "  'PUNCT'],\n",
              " ['VERB',\n",
              "  'PROPN',\n",
              "  'NOUN',\n",
              "  'PUNCT',\n",
              "  'PROPN',\n",
              "  'PROPN',\n",
              "  'PUNCT',\n",
              "  'PRON',\n",
              "  'ADJ',\n",
              "  'NOUN',\n",
              "  'PUNCT',\n",
              "  'SPACE',\n",
              "  'NUM',\n",
              "  'NOUN',\n",
              "  'ADP',\n",
              "  'PROPN',\n",
              "  'PUNCT',\n",
              "  'NUM',\n",
              "  'ADP',\n",
              "  'PROPN',\n",
              "  'PUNCT'],\n",
              " ['ADJ', 'VERB', 'ADP', 'DET', 'NOUN', 'NOUN', 'PRON', 'VERB', 'PROPN'],\n",
              " ['PRON', 'AUX', 'DET', 'NOUN', 'PART', 'VERB', 'NOUN', 'ADV'],\n",
              " ['PUNCT',\n",
              "  'X',\n",
              "  'AUX',\n",
              "  'PRON',\n",
              "  'VERB',\n",
              "  'PRON',\n",
              "  'PROPN',\n",
              "  'PROPN',\n",
              "  'VERB',\n",
              "  'NOUN',\n",
              "  'AUX',\n",
              "  'VERB',\n",
              "  'ADP',\n",
              "  'ADP',\n",
              "  'NOUN',\n",
              "  'PUNCT',\n",
              "  'PRON',\n",
              "  'VERB',\n",
              "  'VERB',\n",
              "  'ADJ',\n",
              "  'ADP',\n",
              "  'DET',\n",
              "  'NOUN',\n",
              "  'PUNCT',\n",
              "  'PROPN'],\n",
              " ['SPACE', 'ADJ', 'NOUN', 'ADP', 'DET', 'PROPN', 'NOUN'],\n",
              " ['PUNCT',\n",
              "  'NUM',\n",
              "  'PUNCT',\n",
              "  'PUNCT',\n",
              "  'PUNCT',\n",
              "  'PROPN',\n",
              "  'PUNCT',\n",
              "  'PROPN',\n",
              "  'PRON',\n",
              "  'ADJ',\n",
              "  'NOUN',\n",
              "  'PUNCT',\n",
              "  'NOUN',\n",
              "  'PUNCT',\n",
              "  'INTJ',\n",
              "  'PUNCT',\n",
              "  'PROPN',\n",
              "  'ADV',\n",
              "  'VERB',\n",
              "  'ADP',\n",
              "  'NOUN',\n",
              "  'NOUN',\n",
              "  'ADV'],\n",
              " ['PUNCT',\n",
              "  'VERB',\n",
              "  'DET',\n",
              "  'NOUN',\n",
              "  'SPACE',\n",
              "  'INTJ',\n",
              "  'ADV',\n",
              "  'PRON',\n",
              "  'PROPN',\n",
              "  'VERB',\n",
              "  'ADV',\n",
              "  'ADJ',\n",
              "  'PART',\n",
              "  'VERB',\n",
              "  'PUNCT'],\n",
              " ['SPACE', 'PROPN', 'PROPN', 'PUNCT', 'ADJ', 'NOUN', 'NOUN'],\n",
              " ['SPACE',\n",
              "  'PRON',\n",
              "  'ADV',\n",
              "  'PUNCT',\n",
              "  'PRON',\n",
              "  'VERB',\n",
              "  'PRON',\n",
              "  'AUX',\n",
              "  'ADV',\n",
              "  'VERB',\n",
              "  'PRON',\n",
              "  'AUX',\n",
              "  'VERB'],\n",
              " ['SPACE', 'PRON', 'VERB', 'ADV', 'ADJ', 'NOUN', 'PUNCT'],\n",
              " ['PRON',\n",
              "  'AUX',\n",
              "  'VERB',\n",
              "  'DET',\n",
              "  'NOUN',\n",
              "  'CCONJ',\n",
              "  'ADP',\n",
              "  'NOUN',\n",
              "  'PRON',\n",
              "  'VERB',\n",
              "  'DET',\n",
              "  'NOUN',\n",
              "  'ADV',\n",
              "  'NUM',\n",
              "  'NOUN',\n",
              "  'ADV',\n",
              "  'ADP',\n",
              "  'NOUN',\n",
              "  'PART',\n",
              "  'NOUN',\n",
              "  'PUNCT',\n",
              "  'PRON',\n",
              "  'VERB',\n",
              "  'ADJ',\n",
              "  'PART',\n",
              "  'VERB',\n",
              "  'DET',\n",
              "  'NOUN',\n",
              "  'PUNCT',\n",
              "  'PRON',\n",
              "  'AUX',\n",
              "  'ADJ',\n",
              "  'PUNCT'],\n",
              " ['AUX',\n",
              "  'PART',\n",
              "  'VERB',\n",
              "  'PRON',\n",
              "  'PART',\n",
              "  'ADV',\n",
              "  'VERB',\n",
              "  'PRON',\n",
              "  'ADP',\n",
              "  'DET',\n",
              "  'NOUN',\n",
              "  'PRON',\n",
              "  'VERB',\n",
              "  'PART',\n",
              "  'VERB'],\n",
              " ['ADJ',\n",
              "  'NOUN',\n",
              "  'PUNCT',\n",
              "  'CCONJ',\n",
              "  'PRON',\n",
              "  'AUX',\n",
              "  'VERB',\n",
              "  'SPACE',\n",
              "  'VERB',\n",
              "  'ADV',\n",
              "  'ADV',\n",
              "  'PRON',\n",
              "  'AUX',\n",
              "  'VERB',\n",
              "  'DET',\n",
              "  'NOUN',\n",
              "  'ADV'],\n",
              " ['CCONJ', 'ADV', 'PRON', 'ADV', 'PRON', 'PUNCT'],\n",
              " ['VERB', 'PART', 'VERB', 'PART', 'VERB', 'ADP', 'PROPN', 'NOUN'],\n",
              " ['PRON',\n",
              "  'AUX',\n",
              "  'PART',\n",
              "  'VERB',\n",
              "  'NOUN',\n",
              "  'PUNCT',\n",
              "  'SCONJ',\n",
              "  'PRON',\n",
              "  'AUX',\n",
              "  'PART',\n",
              "  'VERB',\n",
              "  'DET',\n",
              "  'NOUN',\n",
              "  'PUNCT'],\n",
              " ['SPACE',\n",
              "  'NOUN',\n",
              "  'PUNCT',\n",
              "  'ADV',\n",
              "  'PART',\n",
              "  'PUNCT',\n",
              "  'PRON',\n",
              "  'AUX',\n",
              "  'VERB',\n",
              "  'ADP',\n",
              "  'NUM',\n",
              "  'PUNCT',\n",
              "  'PRON',\n",
              "  'PRON',\n",
              "  'VERB',\n",
              "  'AUX',\n",
              "  'DET',\n",
              "  'NOUN',\n",
              "  'VERB',\n",
              "  'SCONJ',\n",
              "  'PRON',\n",
              "  'VERB',\n",
              "  'ADP',\n",
              "  'PROPN',\n",
              "  'PROPN'],\n",
              " ['ADV',\n",
              "  'VERB',\n",
              "  'ADP',\n",
              "  'NUM',\n",
              "  'NOUN',\n",
              "  'VERB',\n",
              "  'PART',\n",
              "  'VERB',\n",
              "  'ADP',\n",
              "  'PRON',\n",
              "  'NOUN'],\n",
              " ['SPACE',\n",
              "  'INTJ',\n",
              "  'PRON',\n",
              "  'VERB',\n",
              "  'SPACE',\n",
              "  'NOUN',\n",
              "  'AUX',\n",
              "  'AUX',\n",
              "  'ADV',\n",
              "  'ADV',\n",
              "  'ADJ',\n",
              "  'DET',\n",
              "  'NOUN',\n",
              "  'PUNCT',\n",
              "  'NOUN',\n",
              "  'VERB',\n",
              "  'DET',\n",
              "  'NOUN',\n",
              "  'AUX',\n",
              "  'VERB',\n",
              "  'NOUN',\n",
              "  'PUNCT',\n",
              "  'PRON',\n",
              "  'NOUN',\n",
              "  'VERB',\n",
              "  'PRON',\n",
              "  'PUNCT'],\n",
              " ['SPACE',\n",
              "  'PRON',\n",
              "  'VERB',\n",
              "  'DET',\n",
              "  'NOUN',\n",
              "  'PUNCT',\n",
              "  'SCONJ',\n",
              "  'PRON',\n",
              "  'AUX',\n",
              "  'VERB',\n",
              "  'ADV',\n",
              "  'ADJ',\n",
              "  'ADJ',\n",
              "  'NOUN',\n",
              "  'NOUN',\n",
              "  'AUX',\n",
              "  'VERB',\n",
              "  'ADP',\n",
              "  'PRON',\n",
              "  'PUNCT',\n",
              "  'AUX',\n",
              "  'AUX',\n",
              "  'ADV',\n",
              "  'ADJ',\n",
              "  'SPACE'],\n",
              " ['VERB',\n",
              "  'NUM',\n",
              "  'NOUN',\n",
              "  'ADP',\n",
              "  'DET',\n",
              "  'NOUN',\n",
              "  'PUNCT',\n",
              "  'PROPN',\n",
              "  'PROPN',\n",
              "  'NOUN'],\n",
              " ['AUX',\n",
              "  'VERB',\n",
              "  'NOUN',\n",
              "  'ADP',\n",
              "  'DET',\n",
              "  'NOUN',\n",
              "  'ADV',\n",
              "  'ADJ',\n",
              "  'NOUN',\n",
              "  'CCONJ',\n",
              "  'VERB',\n",
              "  'PART',\n",
              "  'VERB',\n",
              "  'DET',\n",
              "  'PROPN',\n",
              "  'NOUN',\n",
              "  'NOUN',\n",
              "  'ADV',\n",
              "  'PUNCT',\n",
              "  'PRON',\n",
              "  'AUX',\n",
              "  'AUX',\n",
              "  'ADJ',\n",
              "  'SCONJ',\n",
              "  'PRON',\n",
              "  'AUX',\n",
              "  'VERB',\n",
              "  'PART',\n",
              "  'VERB',\n",
              "  'DET',\n",
              "  'NOUN',\n",
              "  'CCONJ',\n",
              "  'NOUN',\n",
              "  'CCONJ',\n",
              "  'AUX',\n",
              "  'PRON',\n",
              "  'VERB',\n",
              "  'PART',\n",
              "  'VERB',\n",
              "  'DET',\n",
              "  'NOUN',\n",
              "  'ADP',\n",
              "  'DET',\n",
              "  'NOUN',\n",
              "  'SCONJ',\n",
              "  'PRON',\n",
              "  'VERB',\n",
              "  'SCONJ',\n",
              "  'PRON',\n",
              "  'AUX',\n",
              "  'VERB',\n",
              "  'PART',\n",
              "  'VERB',\n",
              "  'DET',\n",
              "  'NOUN',\n",
              "  'PUNCT',\n",
              "  'VERB',\n",
              "  'DET',\n",
              "  'NOUN',\n",
              "  'CCONJ',\n",
              "  'VERB',\n",
              "  'ADP',\n",
              "  'DET',\n",
              "  'ADJ',\n",
              "  'NOUN',\n",
              "  'PUNCT'],\n",
              " ['PRON',\n",
              "  'ADV',\n",
              "  'VERB',\n",
              "  'DET',\n",
              "  'NOUN',\n",
              "  'VERB',\n",
              "  'ADP',\n",
              "  'PROPN',\n",
              "  'ADP',\n",
              "  'DET',\n",
              "  'NOUN',\n",
              "  'SPACE',\n",
              "  'ADV',\n",
              "  'PRON',\n",
              "  'PRON',\n",
              "  'AUX',\n",
              "  'PRON',\n",
              "  'VERB',\n",
              "  'PART',\n",
              "  'VERB',\n",
              "  'ADP',\n",
              "  'NOUN',\n",
              "  'PUNCT'],\n",
              " ['SPACE',\n",
              "  'VERB',\n",
              "  'PRON',\n",
              "  'NOUN',\n",
              "  'NOUN',\n",
              "  'ADV',\n",
              "  'ADJ',\n",
              "  'PUNCT',\n",
              "  'ADV',\n",
              "  'ADJ',\n",
              "  'CCONJ',\n",
              "  'ADV',\n",
              "  'ADJ',\n",
              "  'PUNCT',\n",
              "  'PRON',\n",
              "  'NOUN'],\n",
              " ['NOUN', 'NOUN', 'NOUN', 'PUNCT'],\n",
              " ['PUNCT',\n",
              "  'PROPN',\n",
              "  'ADJ',\n",
              "  'NOUN',\n",
              "  'PUNCT',\n",
              "  'X',\n",
              "  'ADJ',\n",
              "  'PART',\n",
              "  'VERB',\n",
              "  'PRON',\n",
              "  'PUNCT'],\n",
              " ['SPACE',\n",
              "  'INTJ',\n",
              "  'PRON',\n",
              "  'VERB',\n",
              "  'DET',\n",
              "  'NOUN',\n",
              "  'PUNCT',\n",
              "  'SCONJ',\n",
              "  'PRON',\n",
              "  'AUX',\n",
              "  'ADJ',\n",
              "  'PUNCT'],\n",
              " ['PART',\n",
              "  'ADV',\n",
              "  'ADJ',\n",
              "  'NOUN',\n",
              "  'ADP',\n",
              "  'NOUN',\n",
              "  'NOUN',\n",
              "  'PUNCT',\n",
              "  'AUX',\n",
              "  'AUX',\n",
              "  'ADP',\n",
              "  'DET',\n",
              "  'ADJ',\n",
              "  'NOUN',\n",
              "  'ADP',\n",
              "  'NOUN',\n",
              "  'PUNCT',\n",
              "  'NOUN',\n",
              "  'VERB',\n",
              "  'ADJ',\n",
              "  'ADJ',\n",
              "  'PUNCT',\n",
              "  'ADV',\n",
              "  'ADJ',\n",
              "  'PRON',\n",
              "  'AUX',\n",
              "  'DET',\n",
              "  'ADJ',\n",
              "  'NOUN',\n",
              "  'PUNCT',\n",
              "  'VERB',\n",
              "  'DET',\n",
              "  'NOUN',\n",
              "  'NOUN',\n",
              "  'AUX',\n",
              "  'ADJ',\n",
              "  'NOUN',\n",
              "  'CCONJ',\n",
              "  'ADV',\n",
              "  'VERB',\n",
              "  'PUNCT',\n",
              "  'PROPN',\n",
              "  'ADV',\n",
              "  'VERB',\n",
              "  'PRON',\n",
              "  'VERB',\n",
              "  'PRON',\n",
              "  'ADV',\n",
              "  'ADJ',\n",
              "  'PUNCT',\n",
              "  'VERB',\n",
              "  'NOUN',\n",
              "  'AUX',\n",
              "  'PART',\n",
              "  'PUNCT',\n",
              "  'VERB',\n",
              "  'AUX',\n",
              "  'AUX',\n",
              "  'ADJ',\n",
              "  'PUNCT'],\n",
              " ['SPACE',\n",
              "  'INTJ',\n",
              "  'SPACE',\n",
              "  'PRON',\n",
              "  'VERB',\n",
              "  'ADJ',\n",
              "  'PUNCT',\n",
              "  'PROPN',\n",
              "  'PROPN',\n",
              "  'PROPN',\n",
              "  'ADP',\n",
              "  'NOUN',\n",
              "  'NOUN',\n",
              "  'PUNCT',\n",
              "  'NOUN',\n",
              "  'PUNCT',\n",
              "  'VERB',\n",
              "  'PROPN',\n",
              "  'PROPN'],\n",
              " ['SPACE',\n",
              "  'PROPN',\n",
              "  'PUNCT',\n",
              "  'PROPN',\n",
              "  'PUNCT',\n",
              "  'PUNCT',\n",
              "  'NOUN',\n",
              "  'PROPN',\n",
              "  'PUNCT',\n",
              "  'X',\n",
              "  'X',\n",
              "  'NOUN',\n",
              "  'NOUN',\n",
              "  'PUNCT'],\n",
              " ['ADJ',\n",
              "  'NOUN',\n",
              "  'ADV',\n",
              "  'VERB',\n",
              "  'DET',\n",
              "  'ADJ',\n",
              "  'NOUN',\n",
              "  'PUNCT',\n",
              "  'PRON',\n",
              "  'AUX',\n",
              "  'ADJ',\n",
              "  'PART',\n",
              "  'VERB',\n",
              "  'CCONJ',\n",
              "  'VERB',\n",
              "  'ADJ',\n",
              "  'ADJ',\n",
              "  'NOUN',\n",
              "  'ADP',\n",
              "  'VERB',\n",
              "  'ADJ',\n",
              "  'NOUN',\n",
              "  'PUNCT'],\n",
              " ['VERB',\n",
              "  'DET',\n",
              "  'ADJ',\n",
              "  'NOUN',\n",
              "  'ADV',\n",
              "  'ADV',\n",
              "  'CCONJ',\n",
              "  'ADV',\n",
              "  'PRON',\n",
              "  'VERB',\n",
              "  'NOUN',\n",
              "  'ADP',\n",
              "  'PRON',\n",
              "  'PUNCT'],\n",
              " ['AUX', 'PART', 'VERB', 'ADJ', 'ADP', 'ADV', 'NUM', 'PROPN', 'NOUN'],\n",
              " ['SPACE', 'VERB', 'PRON', 'PUNCT', 'PRON', 'PROPN', 'PROPN'],\n",
              " ['SPACE',\n",
              "  'NOUN',\n",
              "  'PRON',\n",
              "  'VERB',\n",
              "  'DET',\n",
              "  'NOUN',\n",
              "  'CCONJ',\n",
              "  'PRON',\n",
              "  'PRON',\n",
              "  'VERB',\n",
              "  'SCONJ',\n",
              "  'PART',\n",
              "  'VERB',\n",
              "  'PART',\n",
              "  'VERB',\n",
              "  'PROPN',\n",
              "  'PUNCT',\n",
              "  'NOUN',\n",
              "  'PART',\n",
              "  'CCONJ',\n",
              "  'PRON',\n",
              "  'NOUN',\n",
              "  'AUX',\n",
              "  'VERB',\n",
              "  'ADV',\n",
              "  'ADV'],\n",
              " ['PUNCT',\n",
              "  'NOUN',\n",
              "  'NOUN',\n",
              "  'ADJ',\n",
              "  'PRON',\n",
              "  'VERB',\n",
              "  'VERB',\n",
              "  'DET',\n",
              "  'ADJ',\n",
              "  'PUNCT'],\n",
              " ['PRON', 'ADJ', 'NOUN', 'AUX', 'ADV', 'VERB'],\n",
              " ['SPACE',\n",
              "  'DET',\n",
              "  'NOUN',\n",
              "  'NOUN',\n",
              "  'ADP',\n",
              "  'PRON',\n",
              "  'NOUN',\n",
              "  'PUNCT',\n",
              "  'SPACE',\n",
              "  'PROPN',\n",
              "  'VERB',\n",
              "  'VERB',\n",
              "  'PUNCT'],\n",
              " ['PRON',\n",
              "  'DET',\n",
              "  'ADJ',\n",
              "  'NOUN',\n",
              "  'ADP',\n",
              "  'ADJ',\n",
              "  'PUNCT',\n",
              "  'ADV',\n",
              "  'VERB',\n",
              "  'DET',\n",
              "  'NOUN',\n",
              "  'ADP',\n",
              "  'NOUN',\n",
              "  'PRON',\n",
              "  'AUX',\n",
              "  'VERB',\n",
              "  'VERB',\n",
              "  'PRON',\n",
              "  'ADV',\n",
              "  'ADJ',\n",
              "  'PUNCT',\n",
              "  'DET',\n",
              "  'ADJ',\n",
              "  'NOUN',\n",
              "  'AUX',\n",
              "  'PRON',\n",
              "  'VERB',\n",
              "  'NOUN',\n",
              "  'PRON',\n",
              "  'VERB',\n",
              "  'ADV',\n",
              "  'DET',\n",
              "  'NOUN',\n",
              "  'PUNCT',\n",
              "  'PRON',\n",
              "  'AUX',\n",
              "  'VERB',\n",
              "  'PART',\n",
              "  'VERB',\n",
              "  'DET',\n",
              "  'NOUN',\n",
              "  'PART',\n",
              "  'VERB',\n",
              "  'DET',\n",
              "  'NOUN',\n",
              "  'CCONJ',\n",
              "  'INTJ',\n",
              "  'PRON',\n",
              "  'VERB',\n",
              "  'SCONJ',\n",
              "  'ADJ',\n",
              "  'NOUN',\n",
              "  'PUNCT',\n",
              "  'ADV',\n",
              "  'PUNCT',\n",
              "  'SCONJ',\n",
              "  'VERB',\n",
              "  'ADV',\n",
              "  'VERB',\n",
              "  'NOUN',\n",
              "  'PUNCT',\n",
              "  'PRON',\n",
              "  'VERB',\n",
              "  'PART',\n",
              "  'AUX',\n",
              "  'ADJ',\n",
              "  'PART',\n",
              "  'ADV',\n",
              "  'VERB',\n",
              "  'VERB',\n",
              "  'DET',\n",
              "  'NOUN',\n",
              "  'CCONJ',\n",
              "  'VERB',\n",
              "  'PRON',\n",
              "  'VERB',\n",
              "  'DET',\n",
              "  'NOUN',\n",
              "  'NUM',\n",
              "  'ADP',\n",
              "  'VERB',\n",
              "  'DET',\n",
              "  'NOUN',\n",
              "  'CCONJ',\n",
              "  'VERB',\n",
              "  'PART',\n",
              "  'VERB',\n",
              "  'DET',\n",
              "  'NOUN',\n",
              "  'PUNCT',\n",
              "  'ADV',\n",
              "  'PUNCT',\n",
              "  'DET',\n",
              "  'NOUN',\n",
              "  'AUX',\n",
              "  'ADV',\n",
              "  'ADJ',\n",
              "  'PUNCT'],\n",
              " ['PRON',\n",
              "  'AUX',\n",
              "  'VERB',\n",
              "  'PUNCT',\n",
              "  'ADV',\n",
              "  'CCONJ',\n",
              "  'ADV',\n",
              "  'PRON',\n",
              "  'AUX',\n",
              "  'VERB',\n",
              "  'PUNCT',\n",
              "  'PRON',\n",
              "  'VERB',\n",
              "  'VERB',\n",
              "  'PRON',\n",
              "  'NOUN',\n",
              "  'CCONJ',\n",
              "  'PRON',\n",
              "  'AUX',\n",
              "  'PART',\n",
              "  'VERB'],\n",
              " ['PROPN',\n",
              "  'PROPN',\n",
              "  'PUNCT',\n",
              "  'ADV',\n",
              "  'VERB',\n",
              "  'PROPN',\n",
              "  'PROPN',\n",
              "  'PUNCT',\n",
              "  'ADV',\n",
              "  'ADV',\n",
              "  'PART',\n",
              "  'VERB',\n",
              "  'CCONJ',\n",
              "  'VERB',\n",
              "  'ADP',\n",
              "  'PRON',\n",
              "  'ADJ',\n",
              "  'NOUN',\n",
              "  'PUNCT',\n",
              "  'SPACE',\n",
              "  'PRON',\n",
              "  'AUX',\n",
              "  'VERB',\n",
              "  'ADJ',\n",
              "  'NOUN',\n",
              "  'NOUN',\n",
              "  'PUNCT',\n",
              "  'PROPN'],\n",
              " ['SPACE',\n",
              "  'INTJ',\n",
              "  'PUNCT',\n",
              "  'SPACE',\n",
              "  'ADJ',\n",
              "  'NOUN',\n",
              "  'PUNCT',\n",
              "  'PRON',\n",
              "  'VERB',\n",
              "  'ADJ',\n",
              "  'NOUN',\n",
              "  'ADP',\n",
              "  'PRON',\n",
              "  'NOUN'],\n",
              " ['VERB',\n",
              "  'DET',\n",
              "  'ADJ',\n",
              "  'NOUN',\n",
              "  'ADV',\n",
              "  'PRON',\n",
              "  'VERB',\n",
              "  'ADJ',\n",
              "  'ADP',\n",
              "  'NOUN',\n",
              "  'PUNCT',\n",
              "  'PRON',\n",
              "  'NOUN',\n",
              "  'NOUN',\n",
              "  'VERB']]"
            ]
          },
          "metadata": {},
          "execution_count": 41
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "4. Проверим точность разметки. С помощью range выведем для первых пяти текстов пары \"слово - его частеречный тег\"."
      ],
      "metadata": {
        "id": "JECq3nC7f7K6"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Это задание сложнее, чем кажется\n",
        "# Подсказка:\n",
        "# for i in range(len(list)) - это итерация по _длине_ списка;\n",
        "for i in range(5):\n",
        "    doc = nlp(sample[i])\n",
        "    print('Текст', i)\n",
        "    for token in doc:\n",
        "        print(token.text, token.pos_)\n",
        "# в переменную i сохраняется порядковый номер текущего элемента списка;\n",
        "# если у нас есть несколько сопоставимых списков одинаковой длины,\n",
        "# то на каждом шаге итерации мы можем выводить элементы одного порядка из разных списков,\n",
        "# например for i in range(len(list)) поможет вывести tokens[4] и pos[4] одновременно, в одном блоке кода"
      ],
      "metadata": {
        "id": "kZv7_1Ptfxm-",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "9027738e-c446-4703-aa0f-9f1b85934d0c"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Текст 0\n",
            "getting VERB\n",
            "cds NOUN\n",
            "ready ADJ\n",
            "for ADP\n",
            "tour NOUN\n",
            "Текст 1\n",
            "  SPACE\n",
            "MC PROPN\n",
            ", PUNCT\n",
            "happy ADJ\n",
            "mother`s PROPN\n",
            "day NOUN\n",
            "to ADP\n",
            "your PRON\n",
            "mom NOUN\n",
            ";) PUNCT\n",
            ".. PUNCT\n",
            "love NOUN\n",
            "yah NOUN\n",
            "Текст 2\n",
            "A DET\n",
            "year NOUN\n",
            "from ADP\n",
            "now ADV\n",
            "is AUX\n",
            "graduation NOUN\n",
            ".... PUNCT\n",
            "i PRON\n",
            "am AUX\n",
            "pretty ADV\n",
            "sure ADJ\n",
            "i`m PROPN\n",
            "not PART\n",
            "ready ADJ\n",
            "for ADP\n",
            "it PRON\n",
            "! PUNCT\n",
            "? PUNCT\n",
            "! PUNCT\n",
            "? PUNCT\n",
            "! PUNCT\n",
            "? PUNCT\n",
            "Текст 3\n",
            "  SPACE\n",
            "because SCONJ\n",
            "you PRON\n",
            "had VERB\n",
            "chips NOUN\n",
            "and CCONJ\n",
            "sale NOUN\n",
            "w/o ADP\n",
            "me PRON\n",
            "Текст 4\n",
            "Great ADJ\n",
            "for ADP\n",
            "organising VERB\n",
            "my PRON\n",
            "work NOUN\n",
            "life NOUN\n",
            "balance NOUN\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "5. Разметка именованных сущностей\n",
        "\n",
        "Вернемся к списку токенов, который мы создали в пункте 2. Выберем небольшой срез, например тексты 7-19."
      ],
      "metadata": {
        "id": "yLxnLhVgh1ge"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Для каждого текста из выбранного среза: допишите этот код\n",
        "  # Применим к тексту nlp: допишите этот код\n",
        "sample = dataset['test'][7:19]['text']\n",
        "for text in sample:\n",
        "    doc = nlp(text)\n",
        "    print([(ent.text, ent.label_) for ent in doc.ents]) # Вывод именованных сущностей и их лейблов\n"
      ],
      "metadata": {
        "id": "Jw8szOaGgeTe",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "18b1d147-1ae7-4a7d-8481-95eb00bf269a"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[('1st', 'ORDINAL'), ('2890', 'CARDINAL')]\n",
            "[('first', 'ORDINAL')]\n",
            "[('today', 'DATE'), ('i`m', 'NORP')]\n",
            "[]\n",
            "[('Audi', 'ORG')]\n",
            "[]\n",
            "[('REALLY', 'GPE')]\n",
            "[('i`m', 'NORP')]\n",
            "[]\n",
            "[]\n",
            "[('TY', 'ORG')]\n",
            "[('recent years', 'DATE')]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Здесь мы используем новый тип данных - кортеж `(entity, label)`\n",
        "\n",
        "Кортеж очень похож на список. Разница только в том, что кортеж - неизменяемый, т.е. это список, в котором ничего нельзя изменить, удалить или добавить.\n",
        "\n",
        "Кортежи занимают меньше памяти, т.к. у них меньше свойств. Они защищают данные от изменений и могут применяться в сложных структурах, например, служить ключом для словаря `dict`.\n",
        "\n",
        "На практике, их удобно применять для разметки данных  методами NLP."
      ],
      "metadata": {
        "id": "jUAfUQaSlGrM"
      }
    },
    {
      "cell_type": "markdown",
      "source": [],
      "metadata": {
        "id": "Hmm4DP9yn7Ua"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Выберите любой текст и визуализируйте его разметку именованных сущностей\n",
        "sample = dataset['test'][7:23]['text']\n",
        "\n",
        "for text in sample:\n",
        "    doc = nlp(text)\n",
        "    displacy.render(doc, style=\"ent\", jupyter=True)"
      ],
      "metadata": {
        "id": "0aVWyT0gjSoP",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 631
        },
        "outputId": "42c7da0a-ce2a-450a-8bce-9b47f5fc884c"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ],
            "text/html": [
              "<span class=\"tex2jax_ignore\"><div class=\"entities\" style=\"line-height: 2.5; direction: ltr\"> remember the guy who \n",
              "<mark class=\"entity\" style=\"background: #e4e7d2; padding: 0.45em 0.6em; margin: 0 0.25em; line-height: 1; border-radius: 0.35em;\">\n",
              "    1st\n",
              "    <span style=\"font-size: 0.8em; font-weight: bold; line-height: 1; border-radius: 0.35em; vertical-align: middle; margin-left: 0.5rem\">ORDINAL</span>\n",
              "</mark>\n",
              " #tweetbud you! ~&gt; _\n",
              "<mark class=\"entity\" style=\"background: #e4e7d2; padding: 0.45em 0.6em; margin: 0 0.25em; line-height: 1; border-radius: 0.35em;\">\n",
              "    2890\n",
              "    <span style=\"font-size: 0.8em; font-weight: bold; line-height: 1; border-radius: 0.35em; vertical-align: middle; margin-left: 0.5rem\">CARDINAL</span>\n",
              "</mark>\n",
              "  help him get 900 flwrs &amp; make him smile!</div></span>"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ],
            "text/html": [
              "<span class=\"tex2jax_ignore\"><div class=\"entities\" style=\"line-height: 2.5; direction: ltr\"> She! Maybe that was our \n",
              "<mark class=\"entity\" style=\"background: #e4e7d2; padding: 0.45em 0.6em; margin: 0 0.25em; line-height: 1; border-radius: 0.35em;\">\n",
              "    first\n",
              "    <span style=\"font-size: 0.8em; font-weight: bold; line-height: 1; border-radius: 0.35em; vertical-align: middle; margin-left: 0.5rem\">ORDINAL</span>\n",
              "</mark>\n",
              " mistake.  Not everyone is as cool as   (brown nose moment)</div></span>"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ],
            "text/html": [
              "<span class=\"tex2jax_ignore\"><div class=\"entities\" style=\"line-height: 2.5; direction: ltr\">\n",
              "<mark class=\"entity\" style=\"background: #bfe1d9; padding: 0.45em 0.6em; margin: 0 0.25em; line-height: 1; border-radius: 0.35em;\">\n",
              "    today\n",
              "    <span style=\"font-size: 0.8em; font-weight: bold; line-height: 1; border-radius: 0.35em; vertical-align: middle; margin-left: 0.5rem\">DATE</span>\n",
              "</mark>\n",
              " i was so happy got i got off school early, but now \n",
              "<mark class=\"entity\" style=\"background: #c887fb; padding: 0.45em 0.6em; margin: 0 0.25em; line-height: 1; border-radius: 0.35em;\">\n",
              "    i`m\n",
              "    <span style=\"font-size: 0.8em; font-weight: bold; line-height: 1; border-radius: 0.35em; vertical-align: middle; margin-left: 0.5rem\">NORP</span>\n",
              "</mark>\n",
              " bored!!</div></span>"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/spacy/displacy/__init__.py:213: UserWarning: [W006] No entities to visualize found in Doc object. If this is surprising to you, make sure the Doc was processed using a model that supports named entity recognition, and check the `doc.ents` property manually if necessary.\n",
            "  warnings.warn(Warnings.W006)\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ],
            "text/html": [
              "<span class=\"tex2jax_ignore\"><div class=\"entities\" style=\"line-height: 2.5; direction: ltr\"> but but it was funny when I wrote it</div></span>"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ],
            "text/html": [
              "<span class=\"tex2jax_ignore\"><div class=\"entities\" style=\"line-height: 2.5; direction: ltr\"> I saw an all red \n",
              "<mark class=\"entity\" style=\"background: #7aecec; padding: 0.45em 0.6em; margin: 0 0.25em; line-height: 1; border-radius: 0.35em;\">\n",
              "    Audi\n",
              "    <span style=\"font-size: 0.8em; font-weight: bold; line-height: 1; border-radius: 0.35em; vertical-align: middle; margin-left: 0.5rem\">ORG</span>\n",
              "</mark>\n",
              " on the highway. I sped uo to it hoping it was you. My loss</div></span>"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ],
            "text/html": [
              "<span class=\"tex2jax_ignore\"><div class=\"entities\" style=\"line-height: 2.5; direction: ltr\">I loved this app. Worked so great I purchased it. Now it has logged me out and says I don't exist.</div></span>"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ],
            "text/html": [
              "<span class=\"tex2jax_ignore\"><div class=\"entities\" style=\"line-height: 2.5; direction: ltr\">\t\n",
              "<mark class=\"entity\" style=\"background: #feca74; padding: 0.45em 0.6em; margin: 0 0.25em; line-height: 1; border-radius: 0.35em;\">\n",
              "    REALLY\n",
              "    <span style=\"font-size: 0.8em; font-weight: bold; line-height: 1; border-radius: 0.35em; vertical-align: middle; margin-left: 0.5rem\">GPE</span>\n",
              "</mark>\n",
              "?? oh.. sorry yall  lol</div></span>"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ],
            "text/html": [
              "<span class=\"tex2jax_ignore\"><div class=\"entities\" style=\"line-height: 2.5; direction: ltr\"> \n",
              "<mark class=\"entity\" style=\"background: #c887fb; padding: 0.45em 0.6em; margin: 0 0.25em; line-height: 1; border-radius: 0.35em;\">\n",
              "    i`m\n",
              "    <span style=\"font-size: 0.8em; font-weight: bold; line-height: 1; border-radius: 0.35em; vertical-align: middle; margin-left: 0.5rem\">NORP</span>\n",
              "</mark>\n",
              " totally gonna have money to actually buy a pair of shoes this time</div></span>"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ],
            "text/html": [
              "<span class=\"tex2jax_ignore\"><div class=\"entities\" style=\"line-height: 2.5; direction: ltr\">Widget is mostly useless - it constantly displays no items while having a list selected that is well populated. Alternately, the widget becomes logged out somehow. Hope this gets fixed.</div></span>"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ],
            "text/html": [
              "<span class=\"tex2jax_ignore\"><div class=\"entities\" style=\"line-height: 2.5; direction: ltr\">Off to work at the church</div></span>"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ],
            "text/html": [
              "<span class=\"tex2jax_ignore\"><div class=\"entities\" style=\"line-height: 2.5; direction: ltr\">Okay I guess. Very happy with dev response to my previous review, \n",
              "<mark class=\"entity\" style=\"background: #7aecec; padding: 0.45em 0.6em; margin: 0 0.25em; line-height: 1; border-radius: 0.35em;\">\n",
              "    TY\n",
              "    <span style=\"font-size: 0.8em; font-weight: bold; line-height: 1; border-radius: 0.35em; vertical-align: middle; margin-left: 0.5rem\">ORG</span>\n",
              "</mark>\n",
              ".</div></span>"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ],
            "text/html": [
              "<span class=\"tex2jax_ignore\"><div class=\"entities\" style=\"line-height: 2.5; direction: ltr\">  i definitely will , although with the caveat  that i normally only read non fiction..but have more in \n",
              "<mark class=\"entity\" style=\"background: #bfe1d9; padding: 0.45em 0.6em; margin: 0 0.25em; line-height: 1; border-radius: 0.35em;\">\n",
              "    recent years\n",
              "    <span style=\"font-size: 0.8em; font-weight: bold; line-height: 1; border-radius: 0.35em; vertical-align: middle; margin-left: 0.5rem\">DATE</span>\n",
              "</mark>\n",
              " read novels</div></span>"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ],
            "text/html": [
              "<span class=\"tex2jax_ignore\"><div class=\"entities\" style=\"line-height: 2.5; direction: ltr\">Like the look but it messes up straight away. Instead of saving an entry the app cashed and deleted it instead. Don't get!!</div></span>"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ],
            "text/html": [
              "<span class=\"tex2jax_ignore\"><div class=\"entities\" style=\"line-height: 2.5; direction: ltr\">Lots of tasks to complete \n",
              "<mark class=\"entity\" style=\"background: #bfe1d9; padding: 0.45em 0.6em; margin: 0 0.25em; line-height: 1; border-radius: 0.35em;\">\n",
              "    today\n",
              "    <span style=\"font-size: 0.8em; font-weight: bold; line-height: 1; border-radius: 0.35em; vertical-align: middle; margin-left: 0.5rem\">DATE</span>\n",
              "</mark>\n",
              ". \n",
              "<mark class=\"entity\" style=\"background: #e4e7d2; padding: 0.45em 0.6em; margin: 0 0.25em; line-height: 1; border-radius: 0.35em;\">\n",
              "    First\n",
              "    <span style=\"font-size: 0.8em; font-weight: bold; line-height: 1; border-radius: 0.35em; vertical-align: middle; margin-left: 0.5rem\">ORDINAL</span>\n",
              "</mark>\n",
              ", wecloming new members to \n",
              "<mark class=\"entity\" style=\"background: #7aecec; padding: 0.45em 0.6em; margin: 0 0.25em; line-height: 1; border-radius: 0.35em;\">\n",
              "    Canadamigos Social Networking\n",
              "    <span style=\"font-size: 0.8em; font-weight: bold; line-height: 1; border-radius: 0.35em; vertical-align: middle; margin-left: 0.5rem\">ORG</span>\n",
              "</mark>\n",
              " site, www.canadamigos.com, and coffee</div></span>"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ],
            "text/html": [
              "<span class=\"tex2jax_ignore\"><div class=\"entities\" style=\"line-height: 2.5; direction: ltr\"> Oh I believe you...   sooo  thinks she belongs elsewhere, I say I dont think so...</div></span>"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ],
            "text/html": [
              "<span class=\"tex2jax_ignore\"><div class=\"entities\" style=\"line-height: 2.5; direction: ltr\">ughh...i dont feel good</div></span>"
            ]
          },
          "metadata": {}
        }
      ]
    }
  ]
}