{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<img src=\"../data/cosin.png\" width=800>\n",
    "\n",
    "<br></br>\n",
    "# **Chapter 4 | 추천시스템**\n",
    "참고사이트 : https://www.machinelearningplus.com/nlp/cosine-similarity/\n",
    "## **1 데이터 불러오기**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# ! apt-get update\n",
    "# ! apt-get install g++ openjdk-8-jdk \n",
    "# ! pip3  install  nltk konlpy matplotlib gensim \n",
    "\n",
    "# ! apt-get install fonts-nanum-eco\n",
    "# ! apt-get install fontconfig\n",
    "# ! fc-cache -fv\n",
    "# ! cp /usr/share/fonts/truetype/nanum/Nanum* /usr/local/lib/python3.6/dist-packages/matplotlib/mpl-data/fonts/ttf/\n",
    "# ! rm -rf /content/.cache/matplotlib/*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "movies = pd.read_csv('../data/movies_metadata.csv', usecols=['original_title', 'overview', 'title'], low_memory=False)\n",
    "movies = movies.dropna(axis=0)\n",
    "print(movies.shape)\n",
    "\n",
    "movie_plot_li = movies['overview']\n",
    "movie_info_li = movies['title']\n",
    "movies.head(3)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## **2 텍스트 전처리, 모델 만들기**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from nltk.stem import WordNetLemmatizer\n",
    "from nltk.tokenize import RegexpTokenizer\n",
    "\n",
    "class LemmaTokenizer(object):\n",
    "    def __init__(self):\n",
    "        self.wnl       = WordNetLemmatizer()\n",
    "        self.tokenizer = RegexpTokenizer('(?u)[A-z]+')\n",
    "    \n",
    "    def __call__(self, doc):  # 클래스 호출시 마다 실행(Tf-idf Vector 호출)\n",
    "        return([self.wnl.lemmatize(t) for t in self.tokenizer.tokenize(doc)])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# 사이킷런에 위에서 정의한 토크나이저를 입력으로 넣습니다.\n",
    "from sklearn.feature_extraction.text import TfidfVectorizer\n",
    "vectorizer   = TfidfVectorizer(min_df=3, tokenizer=LemmaTokenizer(), \n",
    "                               stop_words='english')\n",
    "X            = vectorizer.fit_transform(movie_plot_li[:10000]) # 메모리 오류로 갯수를 제한\n",
    "vocabluary   = vectorizer.get_feature_names()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# 비슷한 영화 추천하는 Cosin 유사모델 만들기\n",
    "from sklearn.metrics.pairwise import cosine_similarity\n",
    "movie_sim = cosine_similarity(X)\n",
    "print(movie_sim.shape)\n",
    "movie_sim"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## **3 코싸인 유사도 테이블 활용**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# 특정 영화와 유사한 영화목록 출력하기\n",
    "def similar_recommend_by_movie_id(movielens_id, rank=8):\n",
    "    movie_index    = movielens_id - 1\n",
    "    similar_movies = sorted(list(enumerate(movie_sim[movie_index])), key=lambda x:x[1], reverse=True)\n",
    "    print(\"----- {} : 관람객 추천영화 -------\".format(movie_info_li[similar_movies[0][0]]))\n",
    "    for no, movie_idx in enumerate(similar_movies[1:rank]):\n",
    "        print('추천영화 {}순위 : {}'.format(no, movie_info_li[movie_idx[0]]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "similar_recommend_by_movie_id(1, rank=20)"
   ]
  }
 ],
 "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": "3.6.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}