nlp_a2019_tp3/Traitement Articles.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 153,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "from nltk.corpus import stopwords\n",
    "from nltk.tokenize import toktok, sent_tokenize\n",
    "from nltk.parse import CoreNLPParser\n",
    "import re\n",
    "import pickle"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "tok = toktok.ToktokTokenizer()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "pos_tagger = CoreNLPParser(url='http://localhost:9000', tagtype='pos')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "ner_tagger = CoreNLPParser(url='http://localhost:9000', tagtype='ner')\n",
    "#https://github.com/nltk/nltk/wiki/Stanford-CoreNLP-API-in-NLTK"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "textes_articles_df = pd.read_csv(\"refined_data/textes_articles_df.csv\")\n",
    "textes_articles_df = textes_articles_df[textes_articles_df[\"text\"].notnull() & (textes_articles_df[\"media\"]!='CNN')]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 163,
   "metadata": {},
   "outputs": [],
   "source": [
    "del textes_articles_df['Unnamed: 0']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Prétraitement\n",
    "def pretraitement(article):\n",
    "    # tokeniser par phrases\n",
    "    article_sentences = sent_tokenize(article)\n",
    "    article_ner_tokens = []\n",
    "    article_pos_tokens = []\n",
    "    compteur_phrase = 0\n",
    "    for sentence in article_sentences:\n",
    "        # Tokeniser\n",
    "        sentence_tokens = tok.tokenize(sentence)\n",
    "        # Assembler les entités nommées et colocations\n",
    "        sentence_ner = ner_tagger.tag(sentence_tokens)\n",
    "        ner_tokens = [ner_token for ner_token in sentence_ner if ner_token[1] != 'O']\n",
    "        # Supprimer les classes fermées avec un POS\n",
    "        sentence_pos = pos_tagger.tag(sentence_tokens)\n",
    "        pos_tokens = [pos_token for pos_token in sentence_pos if pos_token[1] in ['ADJ','ADV','INTJ','NOUN','PROPN','VERB']]\n",
    "        # Ajouter à la liste de phrases tokenisées\n",
    "        article_ner_tokens.append(ner_tokens)\n",
    "        article_pos_tokens.append(pos_tokens)\n",
    "        compteur_phrase += 1\n",
    "    return article_ner_tokens, article_pos_tokens"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "article_pretraite = [pretraitement(x) for x in list(textes_articles_df[\"text\"])]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 131,
   "metadata": {},
   "outputs": [],
   "source": [
    "def aggreger_ner_tags(article):\n",
    "    dict_named_entity = {}\n",
    "    for sentence in article[0]:\n",
    "        for entity in sentence:\n",
    "            dict_named_entity[entity] = dict_named_entity.get(entity,0) + 1\n",
    "    return dict_named_entity"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 132,
   "metadata": {},
   "outputs": [],
   "source": [
    "def aggreger_pos_tags(article):\n",
    "    dict_pos = {}\n",
    "    for sentence in article[1]:\n",
    "        for pos in sentence:\n",
    "            dict_pos[pos] = dict_pos.get(pos,0) + 1\n",
    "    return dict_pos"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 165,
   "metadata": {},
   "outputs": [],
   "source": [
    "textes_articles_df['ner_dict']=[aggreger_ner_tags(article) for article in article_pretraite]\n",
    "textes_articles_df['pos_dict']=[aggreger_pos_tags(article) for article in article_pretraite]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 167,
   "metadata": {},
   "outputs": [],
   "source": [
    "f = open(\"pickle/textes_articles_df.pickle\",\"wb\")\n",
    "pickle.dump(textes_articles_df,f)\n",
    "f.close()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}