02_Normalisation_des_textes.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "dmNr9PDicuOn"
      },
      "source": [
        "# Normalisation des textes\n",
        "\n",
        "La normalisation d'un texte a pour objectif de diminuer le \"bruit\" et à amplifier ce que l'on cherche à écouter. Par bruit, on entend les variations linguistiques qui sont rares ou non nécessaires à considérer pour la tâche que l'on souhaite réaliser.\n",
        "\n",
        "La **normalisation** d'un texte consiste donc à\n",
        "- supprimer ce qui est rare ou non porteur de sens et qui apporte du bruit,\n",
        "- mettre en avant ce qui est porteur de sens en réduisant/unifiant les variantes.\n",
        "\n",
        "En pratique cela signifie : supprimer les balises  (e.g. _`<div>`_->_∅_), remplacer les variantes d'un mot par une forme référente par expansion systématique des contractions (e.g. _l'_->_le_) ou par lemmatisation ou stemming (e.g. _la_->_le_), supprimer les caractères qui ne sont pas lettres, supprimer les mots vides (e.g. _le_->_∅_), substituer toutes les occurrences d'hashtags par un unique token `#hashtag`, également pour les urls...\n",
        "\n",
        "Suivant la tâche, on peut s'intéroger si la normalisation doit aussi **corriger** les mots mal orthographiés ou les répétitions de caractères qu'ils soient alphabétiques ou autres. En effet la répétition d'une même lettre ou d'un caractère de ponctuation peut marquer l'emphase dans un propos e.g. \"suuuuuuuuuuuper !!!!!\"."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "DUwkUR5ScfiP"
      },
      "source": [
        "## Installation de l'environnement : chargement des modèles et des données\n",
        "\n",
        "Nous utiliserons un corpus de romans de Jules Verne de la littérature française. En pratique le corpus est un répertoire, et les fichiers sont des oeuvres distinctes de Jules Verne. Exécuter le code suivant."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "QBuM_wSby6-5",
        "outputId": "2dd5062e-561c-4fad-ea6f-db8dbe309099",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "# get the corpus and install in local\n",
        "!mkdir data\n",
        "!wget -nc https://raw.githubusercontent.com/nicolashernandez/teaching_nlp/main/data/JulesVerne.zip -P data\n",
        "!unzip data/JulesVerne.zip -d data"
      ],
      "execution_count": 1,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "--2023-09-12 11:11:53--  https://raw.githubusercontent.com/nicolashernandez/teaching_nlp/main/data/JulesVerne.zip\n",
            "Resolving raw.githubusercontent.com (raw.githubusercontent.com)... 185.199.110.133, 185.199.111.133, 185.199.108.133, ...\n",
            "Connecting to raw.githubusercontent.com (raw.githubusercontent.com)|185.199.110.133|:443... connected.\n",
            "HTTP request sent, awaiting response... 200 OK\n",
            "Length: 1822113 (1.7M) [application/zip]\n",
            "Saving to: ‘data/JulesVerne.zip’\n",
            "\n",
            "JulesVerne.zip      100%[===================>]   1.74M  --.-KB/s    in 0.07s   \n",
            "\n",
            "2023-09-12 11:11:54 (26.3 MB/s) - ‘data/JulesVerne.zip’ saved [1822113/1822113]\n",
            "\n",
            "Archive:  data/JulesVerne.zip\n",
            "   creating: data/JulesVerne/\n",
            "  inflating: data/JulesVerne/Jules Verne_Les tribulations d'un chinois en Chine.txt  \n",
            "  inflating: data/JulesVerne/Jules Verne_Voyage au centre de la Terre.txt  \n",
            "  inflating: data/JulesVerne/Jules Verne_20000 Lieues sous les mers.txt  \n",
            "  inflating: data/JulesVerne/Jules Verne_Le tour du monde en quatre-vingts jours.txt  \n",
            "  inflating: data/JulesVerne/Jules Verne_Michel Strogoff.txt  \n",
            "  inflating: data/JulesVerne/Jules Verne_L'île mystérieuse.txt  \n",
            "  inflating: data/JulesVerne/Jules Verne_De la Terre à la Lune.txt  \n",
            "  inflating: data/JulesVerne/Jules Verne_Cinq semaines en ballon.txt  \n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "p_1T3AtFqa2b"
      },
      "source": [
        "Dans le cadre de ce travail nous utiliserons un seul roman (fichier) à savoir _20000 Lieues sous les mers_. Le code suivant permet de vérifier le contenu du fichier ainsi que de définir des variables qui seront utilisées par la suite pour illustrer le fonctionnement des différents codes fournis : `novel_name` et `document`."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "PTQGFlaRsIAG",
        "outputId": "43c0bfaa-c637-4655-dc82-d9f875b8c94b",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "# have a look at a sample of one particular file\n",
        "novel_path = \"data/JulesVerne/Jules Verne_20000 Lieues sous les mers.txt\"\n",
        "\n",
        "import regex as re\n",
        "novel_name = re.search(\"_(.+)\\.txt\", novel_path).group(1)\n",
        "print (novel_name, novel_path)\n",
        "\n",
        "with open(novel_path, encoding='utf8') as f:\n",
        "  document = f.read()\n",
        "  print (document[:1000])"
      ],
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "20000 Lieues sous les mers data/JulesVerne/Jules Verne_20000 Lieues sous les mers.txt\n",
            "20000 Lieues sous les mers\n",
            "\n",
            "                              JULES VERNE\n",
            "                          VINGT MILLE LIEUES\n",
            "                                  SOUS\n",
            "                                LES MERS\n",
            "                              ILLUSTRE DE\n",
            "                      111 DESSINS PAR DE NEUVILLI\n",
            "                              BIBLIOTHEQUE\n",
            "                      D'EDUCATION ET DE RECREATION\n",
            "                     J. HETZEL ET Cie, 18 RUE JACOB\n",
            "                                  PARIS\n",
            "\n",
            "------------------------------------------------------------------------\n",
            "                           TABLE DES MATIÈRES\n",
            "\n",
            "                             PREMIÈRE PARTIE\n",
            "\n",
            "\n",
            "              I       Un écueil fuyant\n",
            "\n",
            "              II      Le pour et le contre\n",
            "\n",
            "              III     Comme il plaira à monsieur\n",
            "\n",
            "              IV      Ned Land\n",
            "\n",
            "              V       À l'aventure !\n",
            "\n",
            "              VI      À toute vapeur\n",
            "\n",
            "              VII     Une baleine d'espèce inconnue\n",
            "\n",
            "              VIII    _Mobilis in mobile_\n",
            "\n",
            "              IX     \n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "EVEwgdpkH6ik"
      },
      "source": [
        "## Mots vides"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "uEHQKPXT3GWW"
      },
      "source": [
        "### QUESTION\n",
        "\n",
        "* Executer le code ci-dessous pour vérifier que la liste de mots vides de nltk pour le français contient bien que des mots vides. Y-a-t-il des mots \"non porteur de sens\" que vous auriez pensé retrouver ?"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Jwl0TDi89v_b",
        "outputId": "f0997716-8103-4072-fb1b-6824bac419c2",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "import nltk\n",
        "nltk.download('stopwords')\n",
        "# retrieve a list of french stop words\n",
        "fr_stop_words = nltk.corpus.stopwords.words('french')\n",
        "print ('fr_stop_words:',fr_stop_words)"
      ],
      "execution_count": 3,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "fr_stop_words: ['au', 'aux', 'avec', 'ce', 'ces', 'dans', 'de', 'des', 'du', 'elle', 'en', 'et', 'eux', 'il', 'ils', 'je', 'la', 'le', 'les', 'leur', 'lui', 'ma', 'mais', 'me', 'même', 'mes', 'moi', 'mon', 'ne', 'nos', 'notre', 'nous', 'on', 'ou', 'par', 'pas', 'pour', 'qu', 'que', 'qui', 'sa', 'se', 'ses', 'son', 'sur', 'ta', 'te', 'tes', 'toi', 'ton', 'tu', 'un', 'une', 'vos', 'votre', 'vous', 'c', 'd', 'j', 'l', 'à', 'm', 'n', 's', 't', 'y', 'été', 'étée', 'étées', 'étés', 'étant', 'étante', 'étants', 'étantes', 'suis', 'es', 'est', 'sommes', 'êtes', 'sont', 'serai', 'seras', 'sera', 'serons', 'serez', 'seront', 'serais', 'serait', 'serions', 'seriez', 'seraient', 'étais', 'était', 'étions', 'étiez', 'étaient', 'fus', 'fut', 'fûmes', 'fûtes', 'furent', 'sois', 'soit', 'soyons', 'soyez', 'soient', 'fusse', 'fusses', 'fût', 'fussions', 'fussiez', 'fussent', 'ayant', 'ayante', 'ayantes', 'ayants', 'eu', 'eue', 'eues', 'eus', 'ai', 'as', 'avons', 'avez', 'ont', 'aurai', 'auras', 'aura', 'aurons', 'aurez', 'auront', 'aurais', 'aurait', 'aurions', 'auriez', 'auraient', 'avais', 'avait', 'avions', 'aviez', 'avaient', 'eut', 'eûmes', 'eûtes', 'eurent', 'aie', 'aies', 'ait', 'ayons', 'ayez', 'aient', 'eusse', 'eusses', 'eût', 'eussions', 'eussiez', 'eussent']\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "[nltk_data] Downloading package stopwords to /root/nltk_data...\n",
            "[nltk_data]   Unzipping corpora/stopwords.zip.\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Yt6AgN8TFWDi"
      },
      "source": [
        "### VOTRE REPONSE\n",
        "\n",
        "**TODO :** Qui ne sont pas dans la liste ?\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "1XAZiKmWIAdw"
      },
      "source": [
        "## Opérations de normalisation"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "SLXCj8w8iypQ"
      },
      "source": [
        "### QUESTION\n",
        "\n",
        "La méthode ci-dessous prétraite un document (une chaîne de caractère) en réalisant un certain nombre d'opérations de normalisation. Replacer les commentaires suivant au bon endroit dans le code.\n",
        "\n",
        "```\n",
        "    # filter tokens with small length\n",
        "    # tokenize document\n",
        "    # filter stopwords out of document\n",
        "    # substitute all characters which are not a whitespace or a letter by a whitespace\n",
        "    # to lower case\n",
        "```\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "FvHJ8ehzbvtk"
      },
      "source": [
        "### VOTRE REPONSE"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "cfRO8m_0dCqB"
      },
      "source": [
        "import nltk\n",
        "\n",
        "# the native re module does not handle \\p{} unicode property\n",
        "import regex as re\n",
        "\n",
        "# define a tokenizer which tokenizes a text by splitting at each whitespace character\n",
        "ws_tokenizer = nltk.WhitespaceTokenizer()\n",
        "\n",
        "# normalize a string of text\n",
        "def normalize_characters(doc):\n",
        "    # substitute all characters which are not a whitespace or a letter by a whitespace\n",
        "    doc = re.sub('[^\\p{L}\\s]', ' ', doc)\n",
        "\n",
        "    # to lower case\n",
        "    doc = doc.lower()\n",
        "\n",
        "    # tokenize document\n",
        "    return ws_tokenizer.tokenize(doc)\n",
        "\n",
        "# normalize tokens\n",
        "def normalize_tokens(tokens):\n",
        "    # filter stopwords out of document\n",
        "    tokens = [token for token in tokens if token not in fr_stop_words]\n",
        "\n",
        "    # filter tokens with small length\n",
        "    tokens = [token for token in tokens if len(token) >2]\n",
        "\n",
        "    # re-create document from filtered tokens\n",
        "    #doc = ' '.join(tokens)\n",
        "    #return doc\n",
        "    return tokens\n",
        "\n",
        "# all in one\n",
        "def preprocess_document(document):\n",
        "  return normalize_tokens(normalize_characters(document))\n"
      ],
      "execution_count": 4,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "RO7sdJ2zAtcf"
      },
      "source": [
        "# le code suivant permet d'illustrer le fonctionnement de la normalisation\n",
        "sample_document = \"Tu pourras (c'est ouf gr!) même goûter au gin de Marina,et mon rhum arrangé!!! 😜🤝🏩🏣👍\"\n",
        "sample_preprocessed_document = normalize_tokens(normalize_characters(sample_document))\n",
        "print ('document:', sample_document)\n",
        "print ('preprocessed_document:', sample_preprocessed_document)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "IDNwUxkZIMUr"
      },
      "source": [
        "## Taille du texte vs. taille de vocabulaire"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "kG1bfSjvwHRU"
      },
      "source": [
        "### QUESTION\n",
        "\n",
        "Le code suivant prétraite le document sans et avec filtrage de tokens (notamment les mots vides). Pour chaque résultat de prétraitement, le nombre total de tokens et la taille du vocabulaire sont affichés.\n",
        "\n",
        "* Comparer les tailles de vocabulaire et le nombre total de tokens avant/après filtrage des mots vides de sens. Que traduisent ces chiffres ?"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "qiVRLLCVstAI",
        "outputId": "c4fb3bf4-d718-403d-bdbd-8bae66e0629b",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "# suppression de certains caractères, passage de tous les caractères en minuscule, tokenization\n",
        "normalized_characters = normalize_characters(document)\n",
        "print ('Before filtering - number of tokens:', len(normalized_characters), '; vocabulary size:', len(set(normalized_characters)))\n",
        "\n",
        "# suppression de certains tokens\n",
        "normalized_tokens = normalize_tokens(normalized_characters)\n",
        "print ('After filtering - number of tokens:', len(normalized_tokens), '; vocabulary size:', len(set(normalized_tokens)))"
      ],
      "execution_count": 5,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Before filtering - number of tokens: 148921 ; vocabulary size: 14639\n",
            "After filtering - number of tokens: 78684 ; vocabulary size: 14458\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "8rrCtcGiw4pY"
      },
      "source": [
        "### VOTRE REPONSE\n",
        "\n",
        "La moitié des tokens composant le document original sont de la ponctuation ou forment des mots supposés non porteurs de sens."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "xjvQz0ILcZhY"
      },
      "source": [
        "## Vérifier si la loi de Zipf s'applique.\n",
        "\n",
        "La loi de zipf :\n",
        "\n",
        "> En classant les mots d’un texte par fréquence décroissante, on observe que la fréquence, `freq(m)`, d’un mot `m` dans un\n",
        "texte est corrélée à son rang `rang(m)` par une\n",
        "loi du type : `freq(m) * rang(m) = K` avec `K`\n",
        "une constante.\n",
        "Peu de mots fréquents et grand nombre de\n",
        "mots rares.\n",
        "\n",
        "La _fréquence d'un mot_ est son _nombre d'occurrence_ sur la somme totale des occurrences de tous les mots. La fréquence constitue une forme de normalisation numérique. Une tendance observée avec le nombre d'occurrences \"devrait\" être la même que celle observée avec une fréquence.\n",
        "\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "q3k7KxBdgbsY"
      },
      "source": [
        "### QUESTION : mots vides vs. mots porteurs de sens\n",
        "\n",
        "Le code ci-dessous traite les mots avant filtrage des mots vides. Il utilise la méthode `FreqDist` d'un paquet de nltk pour calculer la fréquence des tokens du corpus d'un des auteurs, puis affiche les `n` tokens les plus _fréquents_ ; on dit aussi les plus _communs_ (`most_common`).\n",
        "\n",
        "* Exécutez-le. Est-ce bien la fréquence qui est observée ?\n",
        "* Faire varier le nombre de mots les plus fréquents à observer en jouant avec la variable `most_common_n`. Essayer des valeurs de 10 à 1000.  Quand vous regardez les extrêmes (les plus fréquents vs. les moins fréquents des plus fréquents...), que pouvez-vous dire sur la capacité des mots à décrire un contenu (à donner du sens) ?\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "f_LjYwd-fTqO",
        "outputId": "314665ca-2965-487f-ad5f-483b9efa9dcd",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "# import required libraries\n",
        "from nltk.probability import FreqDist\n",
        "\n",
        "# find the frequency\n",
        "fdist = FreqDist(normalized_characters)\n",
        "\n",
        "most_common_n = 1000\n",
        "\n",
        "# print the n most common\n",
        "print ('Les', most_common_n, 'mots les plus fréquents sont :', fdist.most_common(most_common_n))"
      ],
      "execution_count": 8,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Les 1000 mots les plus fréquents sont : [('de', 6316), ('le', 3893), ('et', 3273), ('la', 3222), ('à', 2940), ('l', 2669), ('les', 2592), ('je', 2282), ('des', 2018), ('un', 1822), ('il', 1742), ('d', 1732), ('que', 1609), ('une', 1531), ('en', 1488), ('du', 1428), ('ne', 1300), ('qui', 1279), ('se', 1113), ('nous', 1104), ('ce', 1073), ('dans', 1055), ('pas', 1028), ('au', 950), ('s', 916), ('est', 870), ('par', 833), ('sur', 826), ('capitaine', 751), ('mais', 738), ('plus', 712), ('vous', 712), ('me', 706), ('n', 704), ('ces', 694), ('cette', 663), ('nautilus', 639), ('qu', 630), ('était', 597), ('j', 584), ('pour', 522), ('son', 519), ('a', 491), ('nemo', 485), ('conseil', 481), ('monsieur', 479), ('ned', 454), ('sous', 453), ('c', 451), ('sa', 445), ('avec', 429), ('ses', 412), ('mer', 400), ('comme', 365), ('on', 364), ('si', 359), ('deux', 355), ('m', 351), ('dont', 336), ('lui', 332), ('land', 312), ('avait', 311), ('mon', 310), ('sans', 297), ('quelques', 290), ('leur', 286), ('ou', 284), ('ils', 282), ('dit', 282), ('être', 276), ('moi', 273), ('aux', 269), ('tout', 253), ('même', 252), ('mes', 247), ('bien', 246), ('là', 246), ('y', 243), ('répondit', 236), ('encore', 232), ('canadien', 231), ('où', 228), ('vers', 225), ('eaux', 224), ('cet', 217), ('notre', 214), ('fut', 210), ('sont', 210), ('mètres', 208), ('eau', 200), ('peu', 199), ('donc', 198), ('mille', 195), ('heures', 191), ('alors', 185), ('elle', 185), ('trois', 182), ('après', 180), ('surface', 180), ('très', 178), ('pendant', 175), ('puis', 171), ('entre', 170), ('non', 166), ('ma', 165), ('jusqu', 159), ('cent', 159), ('océan', 159), ('peut', 158), ('fait', 157), ('mers', 155), ('bord', 155), ('leurs', 153), ('ai', 152), ('étaient', 150), ('dire', 147), ('cependant', 144), ('homme', 141), ('ainsi', 141), ('moins', 140), ('cinq', 140), ('quatre', 139), ('quand', 134), ('forme', 132), ('moment', 132), ('vingt', 131), ('ni', 131), ('point', 130), ('air', 130), ('jour', 129), ('milieu', 129), ('rien', 129), ('terre', 128), ('car', 128), ('nos', 126), ('flots', 124), ('jamais', 123), ('dis', 123), ('avoir', 121), ('eût', 121), ('aussi', 120), ('professeur', 120), ('poissons', 120), ('yeux', 118), ('toute', 116), ('temps', 116), ('dix', 114), ('fit', 114), ('effet', 113), ('été', 111), ('fois', 111), ('oui', 111), ('pieds', 110), ('salon', 110), ('main', 109), ('tous', 108), ('milles', 106), ('ont', 104), ('cents', 100), ('toujours', 100), ('avant', 100), ('répondis', 100), ('sud', 100), ('toutes', 99), ('autres', 99), ('animal', 99), ('quelque', 97), ('six', 96), ('autre', 95), ('près', 94), ('commandant', 94), ('tête', 94), ('faut', 93), ('étions', 93), ('seulement', 92), ('huit', 91), ('votre', 91), ('île', 91), ('pouvait', 90), ('elles', 89), ('nord', 88), ('enfin', 88), ('fond', 88), ('ailleurs', 85), ('aronnax', 85), ('depuis', 84), ('nuit', 84), ('corps', 84), ('cela', 84), ('hommes', 84), ('plate', 84), ('avaient', 83), ('navire', 83), ('heure', 82), ('vitesse', 81), ('demi', 81), ('faire', 80), ('plusieurs', 79), ('eh', 79), ('déjà', 78), ('soit', 76), ('profondeur', 76), ('grande', 75), ('doute', 74), ('grand', 74), ('suivant', 74), ('compagnons', 74), ('partie', 73), ('lorsque', 73), ('aucun', 73), ('maître', 73), ('assez', 72), ('or', 72), ('nom', 72), ('devait', 72), ('dessus', 72), ('avez', 72), ('sept', 71), ('t', 71), ('comment', 70), ('avais', 70), ('animaux', 70), ('vis', 70), ('contre', 69), ('monde', 69), ('vie', 68), ('canot', 68), ('jours', 67), ('étais', 67), ('bientôt', 67), ('équipage', 67), ('furent', 66), ('dessous', 65), ('ici', 65), ('ami', 65), ('horizon', 65), ('porte', 65), ('cinquante', 64), ('trente', 64), ('demanda', 64), ('devant', 62), ('nature', 61), ('voir', 61), ('coup', 60), ('seul', 60), ('soleil', 60), ('chambre', 60), ('quel', 58), ('pu', 58), ('frégate', 57), ('trop', 57), ('côte', 57), ('reprit', 57), ('rouge', 57), ('lieues', 56), ('maintenant', 56), ('avions', 56), ('marin', 55), ('latitude', 55), ('ouest', 55), ('ordre', 55), ('bon', 55), ('travers', 55), ('semblait', 54), ('îles', 54), ('matin', 54), ('peine', 54), ('lincoln', 54), ('demandai', 54), ('sol', 54), ('pôle', 54), ('existence', 53), ('abraham', 53), ('hauteur', 53), ('raison', 53), ('celle', 53), ('navires', 52), ('question', 52), ('quelle', 52), ('sommes', 52), ('longues', 52), ('suis', 52), ('esprit', 51), ('ceux', 51), ('large', 51), ('masse', 51), ('marins', 51), ('bateau', 51), ('quoi', 51), ('vos', 51), ('électrique', 51), ('quinze', 50), ('minutes', 50), ('chaque', 50), ('long', 49), ('monstre', 49), ('marche', 49), ('hélice', 49), ('ah', 49), ('appareil', 49), ('prendre', 48), ('cap', 48), ('pression', 48), ('bras', 48), ('écria', 48), ('quarante', 47), ('celui', 47), ('lumière', 47), ('second', 47), ('quant', 46), ('faisait', 46), ('vent', 46), ('arrière', 46), ('compte', 46), ('mieux', 46), ('intérieur', 45), ('fût', 45), ('puisque', 45), ('mois', 45), ('aucune', 45), ('soir', 45), ('pourquoi', 45), ('harponneur', 45), ('millions', 45), ('queue', 45), ('rayons', 45), ('pacifique', 44), ('tant', 44), ('soixante', 44), ('mouvement', 44), ('aurait', 44), ('couches', 44), ('passage', 44), ('sais', 44), ('serait', 44), ('panneaux', 44), ('baleine', 43), ('marine', 43), ('marines', 43), ('parut', 43), ('parmi', 43), ('garçon', 43), ('bruit', 43), ('réservoirs', 43), ('vu', 42), ('fallait', 42), ('longitude', 42), ('route', 42), ('instant', 42), ('harpon', 42), ('environ', 42), ('lendemain', 42), ('pouvais', 42), ('méditerranée', 42), ('côtes', 41), ('atlantique', 41), ('coque', 41), ('laquelle', 41), ('parler', 40), ('situation', 40), ('baleines', 40), ('grands', 40), ('uns', 40), ('vue', 40), ('nouvelle', 40), ('atmosphère', 40), ('tantôt', 40), ('pied', 40), ('détroit', 39), ('pris', 39), ('longs', 39), ('farragut', 39), ('sera', 39), ('globe', 39), ('première', 38), ('immense', 38), ('pus', 38), ('profondeurs', 38), ('sait', 38), ('dos', 38), ('feu', 38), ('espèce', 37), ('eux', 37), ('puissance', 37), ('force', 37), ('eu', 37), ('rapidité', 37), ('nombre', 37), ('véritable', 37), ('mollusques', 37), ('moyen', 37), ('longtemps', 37), ('vapeur', 36), ('haut', 36), ('mot', 36), ('souvent', 36), ('voix', 36), ('crois', 36), ('voilà', 36), ('neuf', 36), ('place', 36), ('ci', 36), ('voici', 36), ('aperçus', 36), ('golfe', 36), ('perles', 36), ('glace', 36), ('vaste', 35), ('abord', 35), ('éclat', 35), ('prit', 35), ('supérieure', 35), ('venait', 35), ('instants', 35), ('aurais', 35), ('libre', 35), ('oiseaux', 35), ('curieux', 35), ('direction', 35), ('banquise', 35), ('présence', 34), ('tard', 34), ('parages', 34), ('croire', 34), ('courant', 34), ('extrême', 34), ('grandes', 34), ('espèces', 34), ('sorte', 34), ('presque', 34), ('allait', 34), ('rapidement', 34), ('oeil', 34), ('répondre', 34), ('droit', 34), ('coeur', 34), ('compris', 34), ('fer', 34), ('travail', 34), ('liquide', 34), ('précisément', 33), ('suivre', 33), ('appareils', 33), ('voulait', 33), ('genre', 33), ('midi', 33), ('table', 32), ('amérique', 32), ('longueur', 32), ('anglais', 32), ('vivement', 32), ('terrible', 32), ('choc', 32), ('lieu', 32), ('conditions', 32), ('côté', 32), ('dents', 32), ('certain', 32), ('cabine', 32), ('mains', 32), ('regards', 32), ('évidemment', 32), ('vint', 32), ('mort', 32), ('nageoires', 32), ('cétacé', 31), ('ligne', 31), ('bonne', 31), ('ans', 31), ('voyage', 31), ('malgré', 31), ('avons', 31), ('ton', 31), ('obscurité', 31), ('liberté', 31), ('bas', 31), ('tour', 30), ('énorme', 30), ('peuvent', 30), ('france', 30), ('formidable', 30), ('cas', 30), ('hasard', 30), ('autour', 30), ('seule', 30), ('écriai', 30), ('nombreux', 30), ('zoophytes', 30), ('température', 30), ('noir', 29), ('corail', 29), ('personne', 29), ('divers', 29), ('mars', 29), ('fort', 29), ('considérable', 29), ('telle', 29), ('mis', 29), ('tôle', 29), ('mots', 29), ('pêche', 29), ('revenir', 29), ('brave', 29), ('observer', 29), ('entendre', 29), ('regard', 29), ('famille', 29), ('loin', 29), ('immobile', 29), ('continents', 28), ('europe', 28), ('chose', 28), ('points', 28), ('vastes', 28), ('dernier', 28), ('narwal', 28), ('certains', 28), ('échantillons', 28), ('journée', 28), ('amis', 28), ('va', 28), ('fus', 28), ('baie', 28), ('lequel', 28), ('pense', 28), ('classe', 28), ('janvier', 28), ('premier', 28), ('entendis', 28), ('panneau', 28), ('fanal', 28), ('février', 28), ('électricité', 27), ('distance', 27), ('donner', 27), ('onze', 27), ('douze', 27), ('plutôt', 27), ('français', 27), ('dernière', 27), ('petit', 27), ('pêcheurs', 27), ('continent', 27), ('sommeil', 27), ('silence', 27), ('lames', 27), ('ciel', 27), ('aller', 27), ('forêts', 27), ('regardai', 27), ('degrés', 27), ('vitre', 27), ('montagne', 27), ('ayant', 26), ('terres', 26), ('histoire', 26), ('mètre', 26), ('tu', 26), ('degré', 26), ('êtes', 26), ('repris', 26), ('tandis', 26), ('demain', 26), ('sentis', 26), ('chair', 26), ('merveilles', 26), ('petites', 26), ('véritables', 26), ('arbres', 26), ('pic', 26), ('promenade', 25), ('pays', 25), ('parfois', 25), ('rapide', 25), ('diverses', 25), ('occasion', 25), ('bâtiment', 25), ('étant', 25), ('petite', 25), ('eut', 25), ('trouvait', 25), ('dû', 25), ('nouveau', 25), ('machine', 25), ('ventre', 25), ('simple', 25), ('fonds', 25), ('cétacés', 25), ('part', 25), ('chasser', 25), ('veux', 25), ('haute', 25), ('double', 25), ('calme', 25), ('donné', 25), ('voyez', 25), ('seconde', 25), ('chasse', 25), ('embarcation', 25), ('paroles', 25), ('élément', 25), ('produits', 25), ('sauvages', 25), ('banc', 25), ('mouvements', 24), ('navigation', 24), ('celles', 24), ('sang', 24), ('subitement', 24), ('chercher', 24), ('plantes', 24), ('fis', 24), ('coquille', 24), ('gros', 24), ('langue', 24), ('parois', 24), ('blanc', 24), ('parce', 24), ('répliqua', 24), ('hautes', 24), ('soudain', 24), ('étrange', 24), ('squales', 24), ('fusil', 24), ('phénomène', 23), ('observations', 23), ('affaire', 23), ('observation', 23), ('donne', 23), ('feux', 23), ('parfaitement', 23), ('absolument', 23), ('voulais', 23), ('pouvaient', 23), ('tenir', 23), ('savez', 23), ('ouvrit', 23), ('montagnes', 23), ('cage', 23), ('profonde', 23), ('vois', 23), ('surpris', 23), ('couleurs', 23), ('quitter', 23), ('petits', 23), ('facile', 23), ('forment', 23), ('forêt', 22), ('année', 22), ('gens', 22), ('connaissance', 22), ('face', 22), ('dernières', 22), ('cri', 22), ('reconnaître', 22), ('naturelle', 22), ('impossible', 22), ('taille', 22), ('éperon', 22), ('voulez', 22), ('allons', 22), ('froid', 22), ('admirable', 22), ('tenait', 22), ('voyais', 22), ('chez', 22), ('aussitôt', 22), ('flancs', 22), ('communication', 22), ('dès', 22), ('spectacle', 22), ('bibliothèque', 22), ('indien', 22), ('sortes', 22), ('plans', 22), ('courants', 22), ('gulf', 22), ('stream', 22), ('archipel', 22), ('flottait', 22), ('glaces', 22), ('particulièrement', 21), ('objet', 21), ('moyenne', 21), ('état', 21), ('opinion', 21), ('couleur', 21), ('régions', 21), ('attention', 21), ('disposition', 21), ('atteindre', 21), ('longue', 21), ('ouverts', 21), ('trouvai', 21), ('faisaient', 21), ('certaine', 21), ('kilogrammes', 21), ('profondes', 21), ('produit', 21), ('manière', 21), ('entendu', 21), ('manomètre', 21), ('coups', 21), ('bouche', 21), ('pensée', 21), ('sentais', 21), ('chaleur', 21), ('semblables', 21), ('remarquai', 21), ('instruments', 21), ('projets', 21), ('urville', 21), ('phoques', 21), ('eussent', 20), ('entier', 20), ('blanche', 20), ('anciens', 20), ('marchait', 20), ('grâce', 20), ('centimètres', 20), ('matelots', 20), ('new', 20), ('départ', 20), ('possible', 20), ('intérêt', 20), ('certainement', 20), ('passe', 20), ('niveau', 20), ('autrefois', 20), ('digne', 20), ('ordres', 20), ('rencontre', 20), ('kilomètres', 20), ('formait', 20), ('goût', 20), ('choses', 20), ('poids', 20), ('savais', 20), ('requins', 20), ('électriques', 20), ('épaisse', 20), ('opération', 20), ('ombre', 20), ('reconnus', 20), ('sommet', 20), ('parole', 20), ('oxygène', 20), ('malheureux', 20), ('filets', 20), ('brun', 20), ('rochers', 20), ('sable', 20), ('raies', 20), ('paris', 19), ('plaine', 19), ('vanikoro', 19), ('tonneaux', 19), ('savoir', 19), ('fin', 19), ('secret', 19), ('pierre', 19), ('voulu', 19), ('derniers', 19), ('laissait', 19), ('masses', 19), ('surtout', 19), ('mit', 19), ('perdre', 19), ('francs', 19), ('saurais', 19), ('geste', 19), ('énormes', 19), ('regardait', 19), ('vrai', 19), ('groupe', 19), ('argent', 19), ('idée', 19), ('lumineux', 19), ('timonier', 19), ('donnait', 19), ('remonter', 19), ('pourrait', 19), ('coquilles', 19), ('aujourd', 19), ('hui', 19), ('observai', 19), ('canal', 19), ('science', 18), ('faites', 18), ('méridien', 18), ('tentacules', 18), ('afin', 18), ('immédiatement', 18), ('mettre', 18), ('produire', 18), ('incessamment', 18), ('noire', 18), ('atmosphères', 18), ('certaines', 18), ('auxquels', 18), ('penser', 18), ('doit', 18), ('prêt', 18), ('compagnon', 18), ('demande', 18), ('crus', 18), ('pouvez', 18), ('cachalots', 18), ('montra', 18), ('passer', 18), ('beau', 18), ('faite', 18), ('signe', 18), ('lignes', 18), ('prison', 18), ('larges', 18), ('paraissait', 18), ('quantité', 18), ('pleine', 18), ('lorsqu', 18), ('franchi', 18), ('etc', 18), ('travaux', 18), ('cuivre', 18), ('éléments', 18), ('algues', 18), ('dumont', 18), ('naturels', 18), ('tunnel', 18), ('poulpes', 18), ('faits', 17), ('livres', 17), ('tel', 17), ('émotion', 17), ('endroit', 17), ('légère', 17), ('êtres', 17), ('gigantesques', 17), ('monstres', 17), ('premiers', 17), ('îlot', 17), ('ouvert', 17), ('ajouta', 17), ('heureusement', 17), ('york', 17), ('veut', 17), ('contraire', 17), ('puissant', 17), ('mammifères', 17), ('pensai', 17), ('arrêta', 17), ('gigantesque', 17), ('trouver', 17), ('ingénieur', 17), ('circonstance', 17), ('seize', 17), ('prenait', 17), ('trompe', 17), ('espace', 17), ('naturaliste', 17), ('existe', 17), ('rendre', 17), ('auquel', 17), ('mêmes', 17), ('besoin', 17), ('dirigea', 17), ('rivages', 17), ('lune', 17), ('disparut', 17), ('riposta', 17), ('fuite', 17), ('évident', 17), ('oreille', 17), ('hors', 17), ('façon', 17), ('laissaient', 17), ('manger', 17), ('règne', 17), ('paraît', 17), ('empêcher', 17), ('murailles', 17), ('connaître', 17), ('mal', 17), ('perle', 17), ('escalier', 17), ('combien', 17), ('excursion', 17), ('suez', 17), ('écueil', 16), ('exactement', 16), ('beaucoup', 16), ('inconnu', 16), ('pouvoir', 16), ('belle', 16), ('font', 16), ('quart', 16), ('put', 16), ('puissante', 16), ('beaux', 16), ('pont', 16), ('trou', 16), ('résultat', 16), ('livre', 16), ('quels', 16), ('ait', 16), ('manquait', 16), ('faisant', 16), ('serons', 16), ('passa', 16), ('élevait', 16), ('rencontrer', 16), ('précieux', 16), ('armés', 16), ('embranchement', 16), ('autant', 16), ('réponse', 16), ('roches', 16), ('extrémité', 16), ('nuages', 16), ('pauvre', 16), ('semblable', 16), ('respiration', 16), ('respirer', 16), ('lueurs', 16), ('position', 16), ('reprendre', 16), ('entendais', 16), ('allais', 16), ('nettement', 16), ('magnifiques', 16), ('portait', 16), ('passé', 16), ('entièrement', 16), ('nombreuses', 16), ('sodium', 16), ('rouges', 16), ('lac', 16), ('rivage', 16), ('dugong', 16), ('fleuve', 15), ('savants', 15), ('existait', 15), ('également', 15), ('carte', 15), ('inférieure', 15), ('guerre', 15), ('danger', 15), ('arrivé', 15), ('naufrage', 15), ('service', 15), ('chacun', 15), ('brise', 15), ('léger', 15), ('prodigieuse', 15), ('prix', 15), ('incident', 15), ('admettre', 15), ('avis', 15), ('variétés', 15), ('atteint', 15), ('naturalistes', 15), ('terribles', 15), ('câble', 15), ('tranquillement', 15), ('laissant', 15), ('lesquelles', 15), ('venir', 15), ('pointe', 15), ('jaune', 15), ('perdu', 15), ('regardant', 15), ('restait', 15), ('sombre', 15), ('légèrement', 15), ('employer', 15), ('violence', 15), ('derrière', 15), ('clarté', 15), ('plein', 15), ('peau', 15), ('poisson', 15), ('emporté', 15), ('cellule', 15), ('apprendre', 15), ('devaient', 15), ('fucus', 15), ('visiter', 15), ('coraux', 15), ('aspect', 15), ('port', 15), ('tombe', 15), ('blocs', 15), ('dangereux', 15), ('vaisseau', 15), ('nageoire', 15), ('mobile', 14), ('torrès', 14), ('officiers', 14), ('toutefois', 14), ('compagnie', 14)]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "QkK6-8i9j_mh"
      },
      "source": [
        "### VOTRE REPONSE\n",
        "\n",
        "- Ce qui est observé est le nombre d'ocurrences.\n",
        "- Les mots les plus fréquents sont les moins porteurs de sens, et inversement ?"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "HNUA8q03wJd6"
      },
      "source": [
        "### QUESTION : observation de la normalisation des mots les plus communs\n",
        "\n",
        "Le code suivant trace une courbe de la \"fréquence\" des mots ordonnée sur le rang de la fréquence.\n",
        "\n",
        "* Dans le code ci-dessous, rajoutez quelques lignes pour visualiser \"aussi\" la \"fréquence\" des mots après normalisation des tokens (i.e. après application de la méthode `normalize_tokens`). Trouvez-vous que la méthode `normalize_tokens` a bien fait son travail ?\n",
        "* Comment caracterisez-vous la 2nd courbe par rapport à la 1ère ?  "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "jx2dLH333xlw",
        "outputId": "457c8b1b-6615-481f-84e5-2d4d261f3966",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        }
      },
      "source": [
        "# plot the graph of frequency for n most common  :\n",
        "import matplotlib.pyplot as plt\n",
        "most_common_n = 40\n",
        "\n",
        "fdist = FreqDist(normalized_characters)\n",
        "print ('Les', most_common_n, 'mots les plus fréquents sont :', fdist.most_common(most_common_n))\n",
        "fdist.plot(most_common_n)\n",
        "\n",
        "fdist = FreqDist(normalized_tokens)\n",
        "print ('Les', most_common_n, 'mots les plus fréquents sont :', fdist.most_common(most_common_n))\n",
        "fdist.plot(most_common_n)"
      ],
      "execution_count": 9,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Les 40 mots les plus fréquents sont : [('de', 6316), ('le', 3893), ('et', 3273), ('la', 3222), ('à', 2940), ('l', 2669), ('les', 2592), ('je', 2282), ('des', 2018), ('un', 1822), ('il', 1742), ('d', 1732), ('que', 1609), ('une', 1531), ('en', 1488), ('du', 1428), ('ne', 1300), ('qui', 1279), ('se', 1113), ('nous', 1104), ('ce', 1073), ('dans', 1055), ('pas', 1028), ('au', 950), ('s', 916), ('est', 870), ('par', 833), ('sur', 826), ('capitaine', 751), ('mais', 738), ('plus', 712), ('vous', 712), ('me', 706), ('n', 704), ('ces', 694), ('cette', 663), ('nautilus', 639), ('qu', 630), ('était', 597), ('j', 584)]\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Les 40 mots les plus fréquents sont : [('capitaine', 751), ('plus', 712), ('cette', 663), ('nautilus', 639), ('nemo', 485), ('conseil', 481), ('monsieur', 479), ('ned', 454), ('sous', 453), ('mer', 400), ('comme', 365), ('deux', 355), ('dont', 336), ('land', 312), ('sans', 297), ('quelques', 290), ('dit', 282), ('être', 276), ('tout', 253), ('bien', 246), ('répondit', 236), ('encore', 232), ('canadien', 231), ('vers', 225), ('eaux', 224), ('cet', 217), ('mètres', 208), ('eau', 200), ('peu', 199), ('donc', 198), ('mille', 195), ('heures', 191), ('alors', 185), ('trois', 182), ('après', 180), ('surface', 180), ('très', 178), ('pendant', 175), ('puis', 171), ('entre', 170)]\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<Axes: xlabel='Samples', ylabel='Counts'>"
            ]
          },
          "metadata": {},
          "execution_count": 9
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "nRJdRmv-z3LG"
      },
      "source": [
        "### VOTRE REPONSE\n",
        "\n",
        "???\n",
        "\n",
        "La loi de Zipf semble toujours s'appliquer, si c'est ce qui est demandé."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "jOZOjGrcb28g"
      },
      "source": [
        "### QUESTION : observation de la fréquence à grande échelle\n",
        "\n",
        "Le code ci-dessous fait globalement la même chose que ci-dessus mais il offre davantage de contrôle sur le calcul de la fréquence. Cela permet aussi de jouer plus finement avec le traçage de la courbe (qui ici permet de considérer tous les mots et non seulement les plus communs).\n",
        "\n",
        "* Executez le code et visualizez la courbe. Eventuellement commenter/décommenter les lignes qui vont bien pour passer à une échelle logarithmique et plus aisément observer des valeurs très petites. Est-ce que la loi de Zipf s'applique sur ce document ?\n",
        "* Remarquez que le code vous permez aussi d'observer la courbe avec le nombre (ou compte) d'occurrences plutôt que la fréquence..."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "wbLplm15V_4T",
        "outputId": "d270cb0b-c2d0-4663-8f33-fbcea19de472",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 476
        }
      },
      "source": [
        "from collections import Counter\n",
        "# We use reduce to concatenate all the lists in tokenized_author_data, but we don't use \"set\"\n",
        "# so that we can count occurencies with a Counter object\n",
        "# Count the number of occurrences of each token\n",
        "occurrences_counter = Counter(normalized_characters)\n",
        "\n",
        "# Formating the counter object to a proper dataset\n",
        "import pandas as pd\n",
        "d = pd.DataFrame(occurrences_counter, index=['occurrences']).transpose().reset_index()\n",
        "d.columns=['word', 'occurrences']\n",
        "\n",
        "# Computing frequencies instead of occurrences\n",
        "nb_total = d.occurrences.sum()\n",
        "d['freq'] = d.occurrences.apply(lambda x: x/nb_total)\n",
        "# d['freq'] = d.occurences.apply(lambda x: x) # <- nombre d'occurrences\n",
        "\n",
        "# Sorting by frequency, most frequent word at the top of the df\n",
        "d = d.sort_values('freq', ascending=False)\n",
        "\n",
        "# Plot\n",
        "import matplotlib.pyplot as plt\n",
        "plt.figure()\n",
        "plt.grid()\n",
        "plt.xscale('log')  # Using log scale\n",
        "plt.yscale('log')  # Using log scale\n",
        "plt.xlabel('log(rank)')\n",
        "plt.ylabel('log(frequency)')\n",
        "#plt.xlabel('rank')\n",
        "#plt.ylabel('frequency')\n",
        "#plt.ylabel('count')\n",
        "plt.title(\"Zipf law : {0}\".format(novel_name))\n",
        "x = list(range(d.shape[0]))\n",
        "plt.plot(x, d.freq)\n",
        "plt.show()"
      ],
      "execution_count": 11,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    }
  ]
}