Semana 7 Modelos Estadísticos.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Semana 7 Modelos Estadísticos",
      "provenance": [],
      "collapsed_sections": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "cWfjPoLSDvN9"
      },
      "source": [
        "# Modelos Estadísticos\n",
        "\n",
        "Algunas de las cosas más interesantes que puedes analizar son específicas al contexto: por ejemplo, si una palabra es un verbo o si un span de texto es el nombre de una persona.\n",
        "\n",
        "Los modelos estadísticos le permiten a spaCy hacer predicciones dentro del contexto. Esto normalmente incluye las part-of-speech tags, dependencias sintácticas y entidades nombradas.\n",
        "\n",
        "Los modelos son entrenados con datasets de textos de ejemplo anotados.\n",
        "\n",
        "Los modelos pueden ser actualizados con más ejemplos para afinar sus predicciones - por ejemplo, para que se desempeñe mejor con tus datos específicos."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "DCBTcU6uEEpP"
      },
      "source": [
        "## Paquetes de modelos\n",
        "\n",
        "spaCy provee varios paquetes de modelos pre-entrenados que puedes descargar usando el comando `spacy download`. Por ejemplo, el paquete `es_core_news_sm` es un modelo pequeño de español que provee soporte para todas las capacidades centrales y está entrenado usando textos de noticias.\n",
        "\n",
        "El método `spacy.load` carga el paquete de modelo por su nombre y devuelve un objeto `nlp`.\n",
        "\n",
        "El paquete provee los parámetros binarios que le permiten a spaCy hacer predicciones.\n",
        "\n",
        "También incluye el vocabulario y la metadata para que spaCy sepa cuál clase de lenguaje usar y cómo configurar el pipeline de procesamiento"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "-Ev_Mo8UEgfC"
      },
      "source": [
        "https://spacy.io/models"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "_vw9CyIIETPT",
        "outputId": "4af19cd4-5b67-4678-8857-4e6bf8852c73"
      },
      "source": [
        "! python -m spacy download es_core_news_sm"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Collecting es_core_news_sm==2.2.5\n",
            "  Downloading https://github.com/explosion/spacy-models/releases/download/es_core_news_sm-2.2.5/es_core_news_sm-2.2.5.tar.gz (16.2 MB)\n",
            "\u001b[K     |████████████████████████████████| 16.2 MB 5.1 MB/s \n",
            "\u001b[?25hRequirement already satisfied: spacy>=2.2.2 in /usr/local/lib/python3.7/dist-packages (from es_core_news_sm==2.2.5) (2.2.4)\n",
            "Requirement already satisfied: blis<0.5.0,>=0.4.0 in /usr/local/lib/python3.7/dist-packages (from spacy>=2.2.2->es_core_news_sm==2.2.5) (0.4.1)\n",
            "Requirement already satisfied: catalogue<1.1.0,>=0.0.7 in /usr/local/lib/python3.7/dist-packages (from spacy>=2.2.2->es_core_news_sm==2.2.5) (1.0.0)\n",
            "Requirement already satisfied: requests<3.0.0,>=2.13.0 in /usr/local/lib/python3.7/dist-packages (from spacy>=2.2.2->es_core_news_sm==2.2.5) (2.23.0)\n",
            "Requirement already satisfied: numpy>=1.15.0 in /usr/local/lib/python3.7/dist-packages (from spacy>=2.2.2->es_core_news_sm==2.2.5) (1.19.5)\n",
            "Requirement already satisfied: cymem<2.1.0,>=2.0.2 in /usr/local/lib/python3.7/dist-packages (from spacy>=2.2.2->es_core_news_sm==2.2.5) (2.0.6)\n",
            "Requirement already satisfied: wasabi<1.1.0,>=0.4.0 in /usr/local/lib/python3.7/dist-packages (from spacy>=2.2.2->es_core_news_sm==2.2.5) (0.8.2)\n",
            "Requirement already satisfied: setuptools in /usr/local/lib/python3.7/dist-packages (from spacy>=2.2.2->es_core_news_sm==2.2.5) (57.4.0)\n",
            "Requirement already satisfied: plac<1.2.0,>=0.9.6 in /usr/local/lib/python3.7/dist-packages (from spacy>=2.2.2->es_core_news_sm==2.2.5) (1.1.3)\n",
            "Requirement already satisfied: srsly<1.1.0,>=1.0.2 in /usr/local/lib/python3.7/dist-packages (from spacy>=2.2.2->es_core_news_sm==2.2.5) (1.0.5)\n",
            "Requirement already satisfied: tqdm<5.0.0,>=4.38.0 in /usr/local/lib/python3.7/dist-packages (from spacy>=2.2.2->es_core_news_sm==2.2.5) (4.62.3)\n",
            "Requirement already satisfied: thinc==7.4.0 in /usr/local/lib/python3.7/dist-packages (from spacy>=2.2.2->es_core_news_sm==2.2.5) (7.4.0)\n",
            "Requirement already satisfied: murmurhash<1.1.0,>=0.28.0 in /usr/local/lib/python3.7/dist-packages (from spacy>=2.2.2->es_core_news_sm==2.2.5) (1.0.6)\n",
            "Requirement already satisfied: preshed<3.1.0,>=3.0.2 in /usr/local/lib/python3.7/dist-packages (from spacy>=2.2.2->es_core_news_sm==2.2.5) (3.0.6)\n",
            "Requirement already satisfied: importlib-metadata>=0.20 in /usr/local/lib/python3.7/dist-packages (from catalogue<1.1.0,>=0.0.7->spacy>=2.2.2->es_core_news_sm==2.2.5) (4.8.2)\n",
            "Requirement already satisfied: zipp>=0.5 in /usr/local/lib/python3.7/dist-packages (from importlib-metadata>=0.20->catalogue<1.1.0,>=0.0.7->spacy>=2.2.2->es_core_news_sm==2.2.5) (3.6.0)\n",
            "Requirement already satisfied: typing-extensions>=3.6.4 in /usr/local/lib/python3.7/dist-packages (from importlib-metadata>=0.20->catalogue<1.1.0,>=0.0.7->spacy>=2.2.2->es_core_news_sm==2.2.5) (3.10.0.2)\n",
            "Requirement already satisfied: certifi>=2017.4.17 in /usr/local/lib/python3.7/dist-packages (from requests<3.0.0,>=2.13.0->spacy>=2.2.2->es_core_news_sm==2.2.5) (2021.10.8)\n",
            "Requirement already satisfied: urllib3!=1.25.0,!=1.25.1,<1.26,>=1.21.1 in /usr/local/lib/python3.7/dist-packages (from requests<3.0.0,>=2.13.0->spacy>=2.2.2->es_core_news_sm==2.2.5) (1.24.3)\n",
            "Requirement already satisfied: chardet<4,>=3.0.2 in /usr/local/lib/python3.7/dist-packages (from requests<3.0.0,>=2.13.0->spacy>=2.2.2->es_core_news_sm==2.2.5) (3.0.4)\n",
            "Requirement already satisfied: idna<3,>=2.5 in /usr/local/lib/python3.7/dist-packages (from requests<3.0.0,>=2.13.0->spacy>=2.2.2->es_core_news_sm==2.2.5) (2.10)\n",
            "Building wheels for collected packages: es-core-news-sm\n",
            "  Building wheel for es-core-news-sm (setup.py) ... \u001b[?25l\u001b[?25hdone\n",
            "  Created wheel for es-core-news-sm: filename=es_core_news_sm-2.2.5-py3-none-any.whl size=16172933 sha256=7d0a3157b6d8a04184e2bf71a654d4efe89a190fa242f7cbc90dd4f921a899c4\n",
            "  Stored in directory: /tmp/pip-ephem-wheel-cache-2vi9mqfd/wheels/21/8d/a9/6c1a2809c55dd22cd9644ae503a52ba6206b04aa57ba83a3d8\n",
            "Successfully built es-core-news-sm\n",
            "Installing collected packages: es-core-news-sm\n",
            "Successfully installed es-core-news-sm-2.2.5\n",
            "\u001b[38;5;2m✔ Download and installation successful\u001b[0m\n",
            "You can now load the model via spacy.load('es_core_news_sm')\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "pMIVuCiSEXbd"
      },
      "source": [
        "import spacy\n",
        "import es_core_news_sm\n",
        "\n",
        "nlp = es_core_news_sm.load()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "HEKlCyXfFAnc"
      },
      "source": [
        "## Prediciendo part-of-speech tags"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "S3KTv-i8FMw4"
      },
      "source": [
        "Revisemos las predicciones del modelo. En este ejemplo estamos usando spaCy para predecir part-of-speech tags, los tipos de palabras en el contexto.\n",
        "\n",
        "Primero, cargamos el modelo pequeño de español y recibimos un objeto nlp.\n",
        "\n",
        "Luego, procesamos el texto \"Ella comió pizza\".\n",
        "\n",
        "Para cada token en el Doc podemos imprimir en pantalla el texto y el part-of-speech tag predicho usando el atributo .pos_.\n",
        "\n",
        "Aquí el modelo predijo correctamente \"comió\" como el verbo y \"pizza\" como el sustantivo."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "UCKT1yPsE86Y",
        "outputId": "50a8e866-5696-4d21-9a25-4cdb40fe56bc"
      },
      "source": [
        "import spacy\n",
        "\n",
        "# Carga el modelo pequeño de español\n",
        "import es_core_news_sm\n",
        "\n",
        "nlp = es_core_news_sm.load()\n",
        "\n",
        "# Procesa un texto\n",
        "doc = nlp(\"Ella comió pizza\")\n",
        "\n",
        "# Itera sobre los tokens\n",
        "for token in doc:\n",
        "    # Imprime en pantalla el texto y el part-of-speech tag predicho\n",
        "    print(token.text, token.pos_)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Ella PRON\n",
            "comió VERB\n",
            "pizza PROPN\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "SBmv6rmyFMUg"
      },
      "source": [
        "## Prediciendo dependencias sintácticas\n",
        "\n",
        "Además de los part-of-speech tags, también podemos predecir las relaciones entre las palabras. Por ejemplo, si una palabra es el sujeto o el objeto de una oración.\n",
        "\n",
        "El atributo `.dep_` devuelve el dependency label predicho.\n",
        "\n",
        "El atributo `.head` devuelve el token cabeza sintáctica. Otra forma de pensarlo es como el token padre al que esta palabra está atada."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "GlCvW_JpFffO",
        "outputId": "0fbcc3dd-c94e-49b1-8395-6571371ca253"
      },
      "source": [
        "for token in doc:\n",
        "    print(token.text, token.pos_, token.dep_, token.head.text)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Ella PRON nsubj comió\n",
            "comió VERB ROOT comió\n",
            "pizza PROPN obj comió\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "25GwQ8O0G7w4"
      },
      "source": [
        "Displacy es un módulo que permite visualizar las dependencias sintácticas."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 246
        },
        "id": "BrH0idVHFzWv",
        "outputId": "495b1684-8d32-4000-ec0a-9a9dbbecef46"
      },
      "source": [
        "from spacy import displacy\n",
        "\n",
        "displacy.render(doc, style='dep', jupyter=True)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "<span class=\"tex2jax_ignore\"><svg xmlns=\"http://www.w3.org/2000/svg\" xmlns:xlink=\"http://www.w3.org/1999/xlink\" xml:lang=\"es\" id=\"94506fbb655449b9b9d4aae3daed13ed-0\" class=\"displacy\" width=\"575\" height=\"224.5\" direction=\"ltr\" style=\"max-width: none; height: 224.5px; color: #000000; background: #ffffff; font-family: Arial; direction: ltr\">\n",
              "<text class=\"displacy-token\" fill=\"currentColor\" text-anchor=\"middle\" y=\"134.5\">\n",
              "    <tspan class=\"displacy-word\" fill=\"currentColor\" x=\"50\">Ella</tspan>\n",
              "    <tspan class=\"displacy-tag\" dy=\"2em\" fill=\"currentColor\" x=\"50\">PRON</tspan>\n",
              "</text>\n",
              "\n",
              "<text class=\"displacy-token\" fill=\"currentColor\" text-anchor=\"middle\" y=\"134.5\">\n",
              "    <tspan class=\"displacy-word\" fill=\"currentColor\" x=\"225\">comió</tspan>\n",
              "    <tspan class=\"displacy-tag\" dy=\"2em\" fill=\"currentColor\" x=\"225\">VERB</tspan>\n",
              "</text>\n",
              "\n",
              "<text class=\"displacy-token\" fill=\"currentColor\" text-anchor=\"middle\" y=\"134.5\">\n",
              "    <tspan class=\"displacy-word\" fill=\"currentColor\" x=\"400\">pizza</tspan>\n",
              "    <tspan class=\"displacy-tag\" dy=\"2em\" fill=\"currentColor\" x=\"400\">PROPN</tspan>\n",
              "</text>\n",
              "\n",
              "<g class=\"displacy-arrow\">\n",
              "    <path class=\"displacy-arc\" id=\"arrow-94506fbb655449b9b9d4aae3daed13ed-0-0\" stroke-width=\"2px\" d=\"M70,89.5 C70,2.0 225.0,2.0 225.0,89.5\" fill=\"none\" stroke=\"currentColor\"/>\n",
              "    <text dy=\"1.25em\" style=\"font-size: 0.8em; letter-spacing: 1px\">\n",
              "        <textPath xlink:href=\"#arrow-94506fbb655449b9b9d4aae3daed13ed-0-0\" class=\"displacy-label\" startOffset=\"50%\" side=\"left\" fill=\"currentColor\" text-anchor=\"middle\">nsubj</textPath>\n",
              "    </text>\n",
              "    <path class=\"displacy-arrowhead\" d=\"M70,91.5 L62,79.5 78,79.5\" fill=\"currentColor\"/>\n",
              "</g>\n",
              "\n",
              "<g class=\"displacy-arrow\">\n",
              "    <path class=\"displacy-arc\" id=\"arrow-94506fbb655449b9b9d4aae3daed13ed-0-1\" stroke-width=\"2px\" d=\"M245,89.5 C245,2.0 400.0,2.0 400.0,89.5\" fill=\"none\" stroke=\"currentColor\"/>\n",
              "    <text dy=\"1.25em\" style=\"font-size: 0.8em; letter-spacing: 1px\">\n",
              "        <textPath xlink:href=\"#arrow-94506fbb655449b9b9d4aae3daed13ed-0-1\" class=\"displacy-label\" startOffset=\"50%\" side=\"left\" fill=\"currentColor\" text-anchor=\"middle\">obj</textPath>\n",
              "    </text>\n",
              "    <path class=\"displacy-arrowhead\" d=\"M400.0,91.5 L408.0,79.5 392.0,79.5\" fill=\"currentColor\"/>\n",
              "</g>\n",
              "</svg></span>"
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "hoik9aDOHGxy"
      },
      "source": [
        "## Prediciendo entidades nombradas\n",
        "\n",
        "Las entidades nombradas son \"objetos de la vida real\" que tienen un nombre asignado. Por ejemplo, una persona, una organización o un país.\n",
        "\n",
        "El atributo `doc.ents` te permite acceder a las entidades nombradas predichas por el modelo.\n",
        "\n",
        "Devuelve un iterador de objetos Span, así que podemos imprimir en pantalla el texto y el label de la entidad usando el atributo `.label_`.\n",
        "\n",
        "En este caso, el modelo predijo correctamente \"Apple\" como una organización, \"EE.UU\" como un lugar, \"iPhone\" y \"Galaxy Note 9\" con la categoría miscelanea."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "E5TzHqvXHWXW",
        "outputId": "185ca1df-38ab-4f26-d868-09daf92f5495"
      },
      "source": [
        "# Procesa un texto\n",
        "doc = nlp(\n",
        "    \"Apple es la marca que más satisfacción genera en EE.UU., \"\n",
        "    \"pero el iPhone, fue superado por el Galaxy Note 9\"\n",
        ")\n",
        "\n",
        "\n",
        "\n",
        "# Itera sobre las entidades predichas\n",
        "for ent in doc.ents:\n",
        "    # Imprime en pantalla el texto y el label de la entidad\n",
        "    print(ent.text, ent.label_)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Apple ORG\n",
            "EE.UU LOC\n",
            "iPhone MISC\n",
            "Galaxy Note 9 MISC\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 35
        },
        "id": "Lx5jYwrfzgLL",
        "outputId": "f8315e98-aaa3-450b-e213-271f3f26eb23"
      },
      "source": [
        "palabra = 'apple iphone'\n",
        "palabra.capitalize()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "string"
            },
            "text/plain": [
              "'Apple iphone'"
            ]
          },
          "metadata": {},
          "execution_count": 12
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "4vKC6CKQO1TC"
      },
      "source": [
        "## Tips: el método `spacy.explain`\n",
        "\n",
        "Para obtener definiciones de los tags y labels más comunes puedes usar la función asistente `spacy.explain`.\n",
        "\n",
        "Por ejemplo, \"LOC\" para lugar no es necesariamente intuitivo, pero spacy.explain puede decirte que se refiere a nombres de ubicaciones definidas política o geográficamente.\n",
        "\n",
        "Lo mismo funciona para part-of-speech tags y dependency labels."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 35
        },
        "id": "LktdzpX2O9u0",
        "outputId": "3cf8159e-3fc3-4962-ad89-97c9c485bc99"
      },
      "source": [
        "spacy.explain(\"LOC\")"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "string"
            },
            "text/plain": [
              "'Non-GPE locations, mountain ranges, bodies of water'"
            ]
          },
          "metadata": {},
          "execution_count": 7
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 35
        },
        "id": "Zn9Fp5D9O_GO",
        "outputId": "f4665788-489b-4fd2-9dc6-7782b557252a"
      },
      "source": [
        "spacy.explain(\"NNP\")"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "string"
            },
            "text/plain": [
              "'noun, proper singular'"
            ]
          },
          "metadata": {},
          "execution_count": 8
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "l_3EKShsPBAV",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 35
        },
        "outputId": "704d79ee-5889-4365-b4b5-a79cac2d1ff6"
      },
      "source": [
        "spacy.explain(\"MISC\")"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "string"
            },
            "text/plain": [
              "'Miscellaneous entities, e.g. events, nationalities, products or works of art'"
            ]
          },
          "metadata": {},
          "execution_count": 9
        }
      ]
    }
  ]
}