Day28- 實作TreeBase模型_作業.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 作業目的: 使用樹型模型進行文章分類\n",
    "\n",
    "本次作業主利用[Amazon Review data中的All Beauty](https://nijianmo.github.io/amazon/index.html)來進行review評價分類(文章分類)\n",
    "\n",
    "資料中將review分為1,2,3,4,5分，而在這份作業，我們將評論改分為差評價、普通評價、優良評價(1,2-->1差評、3-->2普通評價、4,5-->3優良評價)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 載入套件"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import json\n",
    "import re\n",
    "from sklearn.tree import DecisionTreeClassifier\n",
    "from sklearn.ensemble import AdaBoostClassifier, RandomForestClassifier\n",
    "from sklearn.metrics import classification_report, confusion_matrix\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.feature_extraction.text import CountVectorizer, TfidfVectorizer, TfidfTransformer"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 資料前處理\n",
    "文本資料較為龐大，這裡我們取前10000筆資料來進行作業練習"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": [
       "{'overall': 1.0,\n",
       " 'verified': True,\n",
       " 'reviewTime': '02 19, 2015',\n",
       " 'reviewerID': 'A1V6B6TNIC10QE',\n",
       " 'asin': '0143026860',\n",
       " 'reviewerName': 'theodore j bigham',\n",
       " 'reviewText': 'great',\n",
       " 'summary': 'One Star',\n",
       " 'unixReviewTime': 1424304000}"
      ]
     },
     "metadata": {},
     "execution_count": 5
    }
   ],
   "source": [
    "#load json data\n",
    "all_reviews = []\n",
    "###<your code>###\n",
    "with open('All_Beauty.json','r', encoding='utf-8') as f:\n",
    "    for reveiw in f:\n",
    "        all_reviews.append(json.loads(reveiw))\n",
    "\n",
    "\n",
    "all_reviews[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "#parse label(overall) and corpus(reviewText)\n",
    "corpus = []\n",
    "labels = []\n",
    "\n",
    "###<your code>###\n",
    "for review in all_reviews[:10000]:\n",
    "    if 'reviewText' not in review or 'overall' not in reveiw:\n",
    "        continue\n",
    "    corpus.append(review['reviewText'])\n",
    "    labels.append(review['overall'])\n",
    "#transform labels: 1,2 --> 1 and 3 --> 2 and 4,5 --> 3\n",
    "label_map={1:1,2:1,3:2,4:3,5:3}\n",
    "###<your code>###\n",
    "labels= [label_map[int(label)] for label in labels]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "#preprocessing data\n",
    "#remove email address, punctuations, and change line symbol(\\n)\n",
    "\n",
    "###<your code>###\n",
    "pattern = r'\\S*@\\S*|\\\\n|\\W'\n",
    "preprocess_text = lambda x: ' '.join([w for w in re.sub(pattern, ' ', x).split() if w != ''])\n",
    "corpus = [preprocess_text(text) for text in corpus]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": [
       "(8495, 1500, 8495, 1500)"
      ]
     },
     "metadata": {},
     "execution_count": 10
    }
   ],
   "source": [
    "#split corpus and label into train and test\n",
    "###<your code>###\n",
    "x_train, x_test, y_train, y_test = train_test_split(corpus, labels, test_size=0.15, random_state=1)\n",
    "\n",
    "len(x_train), len(x_test), len(y_train), len(y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "#change corpus into vector\n",
    "#you can use tfidf or BoW here\n",
    "###<your code>###\n",
    "vectorizer = TfidfVectorizer()\n",
    "vectorizer.fit(x_train)\n",
    "\n",
    "#transform training and testing corpus into vector form\n",
    "x_train = vectorizer.transform(x_train) ###<your code>###\n",
    "x_test = vectorizer.transform(x_test) ###<your code>###"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 訓練與預測"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": [
       "AdaBoostClassifier(algorithm='SAMME.R',\n",
       "                   base_estimator=DecisionTreeClassifier(ccp_alpha=0.0,\n",
       "                                                         class_weight=None,\n",
       "                                                         criterion='gini',\n",
       "                                                         max_depth=3,\n",
       "                                                         max_features=None,\n",
       "                                                         max_leaf_nodes=None,\n",
       "                                                         min_impurity_decrease=0.0,\n",
       "                                                         min_impurity_split=None,\n",
       "                                                         min_samples_leaf=5,\n",
       "                                                         min_samples_split=10,\n",
       "                                                         min_weight_fraction_leaf=0.0,\n",
       "                                                         presort='deprecated',\n",
       "                                                         random_state=None,\n",
       "                                                         splitter='best'),\n",
       "                   learning_rate=0.8, n_estimators=50, random_state=None)"
      ]
     },
     "metadata": {},
     "execution_count": 12
    }
   ],
   "source": [
    "#build classification model (decision tree, random forest, or adaboost)\n",
    "#start training\n",
    "###<your code>###\n",
    "adaboost_cls = AdaBoostClassifier(base_estimator=DecisionTreeClassifier(criterion='gini',\n",
    "                                                                        max_depth=3,\n",
    "                                                                        min_samples_split=10,\n",
    "                                                                        min_samples_leaf=5),\n",
    "                                  n_estimators=50,\n",
    "                                  learning_rate=0.8)\n",
    "adaboost_cls.fit(x_train,y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "#start inference\n",
    "y_pred = adaboost_cls.predict(x_test) ###<your code>###"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "Accuracy: 0.9046666666666666\n"
     ]
    }
   ],
   "source": [
    "#calculate accuracy\n",
    "###<your code>###\n",
    "print(f\"Accuracy: {adaboost_cls.score(x_test,y_test)}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "              precision    recall  f1-score   support\n\n           1       0.71      0.25      0.38       106\n           2       0.21      0.11      0.14        47\n           3       0.92      0.98      0.95      1347\n\n    accuracy                           0.90      1500\n   macro avg       0.61      0.45      0.49      1500\nweighted avg       0.88      0.90      0.89      1500\n\n[[  27    5   74]\n [   3    5   39]\n [   8   14 1325]]\n"
     ]
    }
   ],
   "source": [
    "#calculate confusion matrix, precision, recall, and f1-score\n",
    "print(classification_report(y_test, y_pred))\n",
    "print(confusion_matrix(y_test, y_pred))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "由上述資訊可以發現, 模型在好評的準確度高(precision, recall都高), 而在差評的部分表現較不理想, 在普通評價的部分大部分跟差評搞混,\n",
    "同學可以試著學習到的各種方法來提升模型的表現"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "name": "python3",
   "display_name": "Python 3.7.6 64-bit (conda)",
   "metadata": {
    "interpreter": {
     "hash": "7b4b4feff2f24a0f0a34464dbe537a36fda679851528fb8735cb41fa49dffb2d"
    }
   }
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.6-final"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}