Simple Classification.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "a1d16ac9",
   "metadata": {},
   "source": [
    "In this notebook I wanna use simple methods such as tfidf or word count. I also use ML models and LSTM."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "771d3bc6",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.feature_extraction.text import CountVectorizer\n",
    "from sklearn.feature_extraction.text import TfidfVectorizer\n",
    "from sklearn.naive_bayes import MultinomialNB\n",
    "from sklearn.tree import DecisionTreeClassifier\n",
    "from sklearn.svm import SVC\n",
    "import tensorflow as tf\n",
    "from tensorflow.keras.preprocessing.text import Tokenizer\n",
    "from tensorflow.keras.preprocessing.sequence import pad_sequences\n",
    "from tensorflow.keras.models import Sequential\n",
    "from tensorflow.keras.layers import Embedding, LSTM, Dense, Dropout\n",
    "from tensorflow.keras.optimizers import Adam\n",
    "from tensorflow.keras import layers\n",
    "from sklearn.preprocessing import LabelBinarizer\n",
    "from sklearn.metrics import classification_report, confusion_matrix\n",
    "\n",
    "import warnings\n",
    "warnings.filterwarnings(action='ignore')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "52136d4f",
   "metadata": {},
   "outputs": [],
   "source": [
    "df=pd.read_csv(r'final_train.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "5711c149",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>text</th>\n",
       "      <th>class</th>\n",
       "      <th>PP_Text_2</th>\n",
       "      <th>length</th>\n",
       "      <th>num_tokens</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>\\nاز سوي كميسيون ماده 10 \\nاحزابكارگزاران سازن...</td>\n",
       "      <td>Politics</td>\n",
       "      <td>کمیسیون ماده ۱۰ احزابکارگزاران سازندگی مجوز فع...</td>\n",
       "      <td>344</td>\n",
       "      <td>56</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>\\nخصوصي سازي كارخانه هاي چاي سال آينده \\nآغاز ...</td>\n",
       "      <td>Economy</td>\n",
       "      <td>خصوصی کارخانه چای آینده آغاز مدیرعامل سازمان چ...</td>\n",
       "      <td>687</td>\n",
       "      <td>113</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>\\nتوضيح يك مطلب \\nسوم بهمن ماه امسال گزارش طرح...</td>\n",
       "      <td>Science and Culture</td>\n",
       "      <td>توضیح مطلب بهمن ماه امسال گزارش طرح بررسی شناس...</td>\n",
       "      <td>1120</td>\n",
       "      <td>196</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>\\nبهاي گازوئيل و بنزين در \\nفرانسه افزايش يافت...</td>\n",
       "      <td>Economy</td>\n",
       "      <td>بهای گازوئیل بنزین فرانسه افزایش پاریس خبرگزار...</td>\n",
       "      <td>396</td>\n",
       "      <td>70</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>\\nرئيس سازمان حج و زيارت اعلام كرد \\n شهريور 1...</td>\n",
       "      <td>Science and Culture</td>\n",
       "      <td>رئیس سازمان حج زیارت اعلام شهریور ۱۶ زمان ثبت ...</td>\n",
       "      <td>495</td>\n",
       "      <td>95</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                text                class  \\\n",
       "0  \\nاز سوي كميسيون ماده 10 \\nاحزابكارگزاران سازن...             Politics   \n",
       "1  \\nخصوصي سازي كارخانه هاي چاي سال آينده \\nآغاز ...              Economy   \n",
       "2  \\nتوضيح يك مطلب \\nسوم بهمن ماه امسال گزارش طرح...  Science and Culture   \n",
       "3  \\nبهاي گازوئيل و بنزين در \\nفرانسه افزايش يافت...              Economy   \n",
       "4  \\nرئيس سازمان حج و زيارت اعلام كرد \\n شهريور 1...  Science and Culture   \n",
       "\n",
       "                                           PP_Text_2  length  num_tokens  \n",
       "0  کمیسیون ماده ۱۰ احزابکارگزاران سازندگی مجوز فع...     344          56  \n",
       "1  خصوصی کارخانه چای آینده آغاز مدیرعامل سازمان چ...     687         113  \n",
       "2  توضیح مطلب بهمن ماه امسال گزارش طرح بررسی شناس...    1120         196  \n",
       "3  بهای گازوئیل بنزین فرانسه افزایش پاریس خبرگزار...     396          70  \n",
       "4  رئیس سازمان حج زیارت اعلام شهریور ۱۶ زمان ثبت ...     495          95  "
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "af362773",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'جلسه علنی دیروز اصلاحیه قانون صدا سیما مسکوت مان طرح جامع رئیس مجلس طرح تحقیق تفحص صداوسیما امروز مجلس مطرح گروه سیاسی طرح الحاق ماده تبصره قانون اساسنامه سازمان صداوسیما جمهوری اسلام ایران دیروز جلسه علنی مجلس درخواست کمیسیون فرهنگ تصویب نمایندگان مجلس شورا اسلام ماه مسکوت مان فوریت طرح امضا ۲۵ تن نمایندگان رسیده نوزدهم فروردین ماه قید فوریت تقدیم مجلس فوریت تصویب نمایندگان ماده امضاءکنندگان طرح خواستار افزودن قانون اساسنامه سازمان صداوسیما جمهوری اسلام ایران مصوب ۱۳۶۲ چنانچه برنامه صداوسیما اعم تولید گزارش خبر قالب مختلف تصویر نمایش شبکه محل المللی اطلاعیه صادره سازمان مطالب مشتمل توهین افترا خلاف واقع اعم حقیقی یاحقوقی هرنحوی اظهارات اشخاص تحریف درخواست ذینفع سازمان صداوسیما مکلف توضیحات پاسخ ذینفع مشروط برآنکه متضمن توهین افترا تاحداکثر دوبرابرمدت زمان مطلب اصل حداکثرساعت ۲۴ دریافت جوابیه دره زمان برنامه اولیه مجانی خواست مسکوت ماندن طرح کمیسیون فرهنگ ارائه طرح کمیسیون قانون اداره صدا سیما جلسه علنی دیروز مجلس لایحه اعاده شورا نگهبان مطرح اصلاح آنهابرای تامین شورا عمل لایحه موافقتنامه دولت جمهوری اسلام ایران دولت فدراسیون روسیه جمهوری اسلام پاکستان جمهوری عرب سوریه اوکراین جمهوری سوسیالیستی دموکراتیک سریلانکا وقطر اجتناب اخذ مالیات مضاعف جلوگیری فرار مالی مالیات درآمد سرمایه تقدیم مجلس شورا اسلام درادامه جلسه لایحه الحاق دولت جمهوری اسلام ایران کنوانسیون سازمان کنفرانس اسلام جهت مبارزه تروریسم الملل شورا نگهبان بازگشت مطرح توجه اصرار مجلس مصوبه لایحه تکلیف نهایی مجمع تشخیص مصلحت نظام فرست دراین جلسه کلیات لایحه اصلاح مواد قانون مالیات مستقیم بررسی تصویب نمایندگان مجلس شورا اسلام گذر مجلس شورا اسلام امروز جلسه علنی دیروز رئیس مجلس شورا اسلام اعلام طرح تحقیق تفحص صدا سیما جلسه علنی روز شنبه مجلس مطرح کروبی پاسخ تذکرآیین نامه هادی نماینده مردم محلات دلیجان درمورد مطرح نشدن طرح تحقیق تفحص صدا سیما اظهار دلیل اصل مطرح نشدن طرح تعطیلیهای رسم'"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['PP_Text_2'][1684]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "69dd9863",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "text           0\n",
       "class          0\n",
       "PP_Text_2     11\n",
       "length         0\n",
       "num_tokens     0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.isna().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "d1906ce0",
   "metadata": {},
   "outputs": [],
   "source": [
    "hf=df[df['PP_Text_2'].isnull()==True]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "472981a9",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>text</th>\n",
       "      <th>class</th>\n",
       "      <th>PP_Text_2</th>\n",
       "      <th>length</th>\n",
       "      <th>num_tokens</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>16896</th>\n",
       "      <td>\\nاست. \\n</td>\n",
       "      <td>Science and Culture</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23627</th>\n",
       "      <td>\\nشد. \\n</td>\n",
       "      <td>Economy</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24823</th>\n",
       "      <td>\\nاست. \\n</td>\n",
       "      <td>Social</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27631</th>\n",
       "      <td>\\nبود. \\n</td>\n",
       "      <td>Sport</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28157</th>\n",
       "      <td>\\nبود. \\n</td>\n",
       "      <td>Politics</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29374</th>\n",
       "      <td>\\nمي گرفت. \\n</td>\n",
       "      <td>Economy</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29456</th>\n",
       "      <td>\\nدارد! \\n</td>\n",
       "      <td>Sport</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>30025</th>\n",
       "      <td>\\nاست. \\n</td>\n",
       "      <td>Economy</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32949</th>\n",
       "      <td>\\n؟ باشند\\n</td>\n",
       "      <td>Sport</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41194</th>\n",
       "      <td>\\nنداشت!!! \\n</td>\n",
       "      <td>Sport</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>56461</th>\n",
       "      <td>\\nبود. \\n</td>\n",
       "      <td>Sport</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                text                class PP_Text_2  length  num_tokens\n",
       "16896      \\nاست. \\n  Science and Culture       NaN       0           0\n",
       "23627       \\nشد. \\n              Economy       NaN       0           0\n",
       "24823      \\nاست. \\n               Social       NaN       0           0\n",
       "27631      \\nبود. \\n                Sport       NaN       0           0\n",
       "28157      \\nبود. \\n             Politics       NaN       0           0\n",
       "29374  \\nمي گرفت. \\n              Economy       NaN       0           0\n",
       "29456     \\nدارد! \\n                Sport       NaN       0           0\n",
       "30025      \\nاست. \\n              Economy       NaN       0           0\n",
       "32949    \\n؟ باشند\\n                Sport       NaN       0           0\n",
       "41194  \\nنداشت!!! \\n                Sport       NaN       0           0\n",
       "56461      \\nبود. \\n                Sport       NaN       0           0"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "hf"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "a05d6e3e",
   "metadata": {},
   "outputs": [],
   "source": [
    "df.dropna(axis=0,inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "0c821346",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "7a5234c3",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "a37238c1",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "id": "529ac130",
   "metadata": {},
   "source": [
    "# CountVectorizer Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "b928e974",
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train, X_val, y_train, y_val = train_test_split(df['PP_Text_2'], df['class'], test_size=0.2, random_state=42)\n",
    "\n",
    "\n",
    "count_vec = CountVectorizer()\n",
    "count_vec_train = count_vec.fit_transform(X_train)\n",
    "count_vec_val = count_vec.transform(X_val)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "f210e70c",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train_X shape:  (47982, 210205)\n",
      "Validation_X shape:  (11996, 210205)\n"
     ]
    }
   ],
   "source": [
    "print(\"Train_X shape: \",count_vec_train.shape)\n",
    "print(\"Validation_X shape: \",count_vec_val.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "f39e2635",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Classification Report:\n",
      "                     precision    recall  f1-score   support\n",
      "\n",
      "            Economy       0.92      0.91      0.91      3173\n",
      "           Politics       0.83      0.90      0.87      3067\n",
      "Science and Culture       0.82      0.84      0.83      1814\n",
      "             Social       0.80      0.69      0.74      1422\n",
      "              Sport       0.99      0.97      0.98      2520\n",
      "\n",
      "           accuracy                           0.88     11996\n",
      "          macro avg       0.87      0.86      0.87     11996\n",
      "       weighted avg       0.88      0.88      0.88     11996\n",
      "\n",
      "Confusion Matrix:\n",
      " [[2872  187   61   43   10]\n",
      " [ 135 2760   81   80   11]\n",
      " [  48  141 1516  106    3]\n",
      " [  72  208  153  984    5]\n",
      " [   5   14   32   17 2452]]\n"
     ]
    }
   ],
   "source": [
    "classifier = MultinomialNB()\n",
    "classifier.fit(count_vec_train, y_train)\n",
    "\n",
    "y_pred = classifier.predict(count_vec_val)\n",
    "\n",
    "# Evaluate the model\n",
    "print(\"Classification Report:\")\n",
    "print(classification_report(y_val, y_pred))\n",
    "\n",
    "# Generate confusion matrix\n",
    "confusion = confusion_matrix(y_val, y_pred)\n",
    "print(\"Confusion Matrix:\\n\", confusion)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "555269f3",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Classification Report:\n",
      "                     precision    recall  f1-score   support\n",
      "\n",
      "            Economy       0.85      0.87      0.86      3173\n",
      "           Politics       0.83      0.83      0.83      3067\n",
      "Science and Culture       0.77      0.77      0.77      1814\n",
      "             Social       0.69      0.66      0.67      1422\n",
      "              Sport       0.97      0.97      0.97      2520\n",
      "\n",
      "           accuracy                           0.84     11996\n",
      "          macro avg       0.82      0.82      0.82     11996\n",
      "       weighted avg       0.84      0.84      0.84     11996\n",
      "\n",
      "Confusion Matrix:\n",
      " [[2763  226   83   89   12]\n",
      " [ 253 2551  101  136   26]\n",
      " [ 101  122 1397  170   24]\n",
      " [ 110  142  215  932   23]\n",
      " [  14   22   26   17 2441]]\n"
     ]
    }
   ],
   "source": [
    "DT=DecisionTreeClassifier()\n",
    "DT.fit(count_vec_train, y_train)\n",
    "\n",
    "y_pred_2 = DT.predict(count_vec_val)\n",
    "\n",
    "# Evaluate the model\n",
    "print(\"Classification Report:\")\n",
    "print(classification_report(y_val, y_pred_2))\n",
    "\n",
    "# Generate confusion matrix\n",
    "confusion = confusion_matrix(y_val, y_pred_2)\n",
    "print(\"Confusion Matrix:\\n\", confusion)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "decb9502",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Classification Report:\n",
      "                     precision    recall  f1-score   support\n",
      "\n",
      "            Economy       0.91      0.93      0.92      3173\n",
      "           Politics       0.88      0.89      0.88      3067\n",
      "Science and Culture       0.85      0.83      0.84      1814\n",
      "             Social       0.78      0.75      0.76      1422\n",
      "              Sport       0.98      0.98      0.98      2520\n",
      "\n",
      "           accuracy                           0.89     11996\n",
      "          macro avg       0.88      0.87      0.88     11996\n",
      "       weighted avg       0.89      0.89      0.89     11996\n",
      "\n",
      "Confusion Matrix:\n",
      " [[2953  132   40   44    4]\n",
      " [ 158 2724   62  111   12]\n",
      " [  47  101 1508  146   12]\n",
      " [  76  139  131 1065   11]\n",
      " [  13   16   27    5 2459]]\n"
     ]
    }
   ],
   "source": [
    "svm = SVC(kernel='linear')  \n",
    "\n",
    "\n",
    "svm.fit(count_vec_train, y_train)\n",
    "y_pred_3 = svm.predict(count_vec_val)\n",
    "\n",
    "print(\"Classification Report:\")\n",
    "print(classification_report(y_val, y_pred_3))\n",
    "\n",
    "# Generate confusion matrix\n",
    "confusion = confusion_matrix(y_val, y_pred_3)\n",
    "print(\"Confusion Matrix:\\n\", confusion)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "51f0250a",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "4e7e7e93",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "2b664d03",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "faf09385",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "id": "945611be",
   "metadata": {},
   "source": [
    "# Tfidf Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "bd6c331b",
   "metadata": {},
   "outputs": [],
   "source": [
    "tfidf = TfidfVectorizer()\n",
    "tfidf_train = tfidf.fit_transform(X_train)\n",
    "tfidf_val = tfidf.transform(X_val)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "03c18cd6",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Tfidf Train_X shape:  (47982, 210205)\n",
      "Tfidf Validation_X shape:  (11996, 210205)\n"
     ]
    }
   ],
   "source": [
    "print(\"Tfidf Train_X shape: \",tfidf_train.shape)\n",
    "print(\"Tfidf Validation_X shape: \",tfidf_val.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "fbf1cc8c",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Classification Report:\n",
      "                     precision    recall  f1-score   support\n",
      "\n",
      "            Economy       0.87      0.91      0.89      3173\n",
      "           Politics       0.67      0.94      0.78      3067\n",
      "Science and Culture       0.83      0.71      0.77      1814\n",
      "             Social       0.91      0.22      0.35      1422\n",
      "              Sport       0.99      0.96      0.98      2520\n",
      "\n",
      "           accuracy                           0.82     11996\n",
      "          macro avg       0.85      0.75      0.75     11996\n",
      "       weighted avg       0.84      0.82      0.80     11996\n",
      "\n",
      "Confusion Matrix:\n",
      " [[2899  244   23    1    6]\n",
      " [ 143 2873   38    1   12]\n",
      " [ 116  374 1293   26    5]\n",
      " [ 158  772  174  310    8]\n",
      " [  13   49   26    1 2431]]\n"
     ]
    }
   ],
   "source": [
    "classifier = MultinomialNB()\n",
    "classifier.fit(tfidf_train, y_train)\n",
    "\n",
    "y_pred = classifier.predict(tfidf_val)\n",
    "\n",
    "# Evaluate the model\n",
    "print(\"Classification Report:\")\n",
    "print(classification_report(y_val, y_pred))\n",
    "\n",
    "# Generate confusion matrix\n",
    "confusion = confusion_matrix(y_val, y_pred)\n",
    "print(\"Confusion Matrix:\\n\", confusion)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "056fd04c",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Classification Report:\n",
      "                     precision    recall  f1-score   support\n",
      "\n",
      "            Economy       0.86      0.87      0.87      3173\n",
      "           Politics       0.82      0.82      0.82      3067\n",
      "Science and Culture       0.76      0.78      0.77      1814\n",
      "             Social       0.69      0.67      0.68      1422\n",
      "              Sport       0.97      0.97      0.97      2520\n",
      "\n",
      "           accuracy                           0.84     11996\n",
      "          macro avg       0.82      0.82      0.82     11996\n",
      "       weighted avg       0.84      0.84      0.84     11996\n",
      "\n",
      "Confusion Matrix:\n",
      " [[2754  239   90   78   12]\n",
      " [ 241 2526  124  158   18]\n",
      " [  77  135 1406  168   28]\n",
      " [  97  160  193  947   25]\n",
      " [  17   14   31   25 2433]]\n"
     ]
    }
   ],
   "source": [
    "DT=DecisionTreeClassifier()\n",
    "DT.fit(tfidf_train, y_train)\n",
    "\n",
    "y_pred_2 = DT.predict(tfidf_val)\n",
    "\n",
    "# Evaluate the model\n",
    "print(\"Classification Report:\")\n",
    "print(classification_report(y_val, y_pred_2))\n",
    "\n",
    "# Generate confusion matrix\n",
    "confusion = confusion_matrix(y_val, y_pred_2)\n",
    "print(\"Confusion Matrix:\\n\", confusion)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "993ed0fa",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Classification Report:\n",
      "                     precision    recall  f1-score   support\n",
      "\n",
      "            Economy       0.94      0.96      0.95      3173\n",
      "           Politics       0.92      0.92      0.92      3067\n",
      "Science and Culture       0.88      0.90      0.89      1814\n",
      "             Social       0.86      0.81      0.83      1422\n",
      "              Sport       0.99      0.99      0.99      2520\n",
      "\n",
      "           accuracy                           0.93     11996\n",
      "          macro avg       0.92      0.91      0.92     11996\n",
      "       weighted avg       0.93      0.93      0.93     11996\n",
      "\n",
      "Confusion Matrix:\n",
      " [[3034   68   43   23    5]\n",
      " [ 108 2811   57   78   13]\n",
      " [  30   50 1637   89    8]\n",
      " [  50  106  110 1151    5]\n",
      " [   3    4   17    5 2491]]\n"
     ]
    }
   ],
   "source": [
    "svm = SVC(kernel='linear')  \n",
    "\n",
    "\n",
    "svm.fit(tfidf_train, y_train)\n",
    "y_pred_3 = svm.predict(tfidf_val)\n",
    "\n",
    "print(\"Classification Report:\")\n",
    "print(classification_report(y_val, y_pred_3))\n",
    "\n",
    "# Generate confusion matrix\n",
    "confusion = confusion_matrix(y_val, y_pred_3)\n",
    "print(\"Confusion Matrix:\\n\", confusion)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "fb397342",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "b2fa0a8d",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "20b04b8c",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "id": "ac196be4",
   "metadata": {},
   "source": [
    "# LSTM And Word Embedding"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "5886c5ff",
   "metadata": {},
   "outputs": [],
   "source": [
    "X = df['PP_Text_2']\n",
    "tokenizer = Tokenizer(num_words=20000)\n",
    "tokenizer.fit_on_texts(X)\n",
    "X=tokenizer.texts_to_sequences(X)\n",
    "X=pad_sequences(X,maxlen=100)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "8207cafc",
   "metadata": {},
   "outputs": [],
   "source": [
    "labels = df['class'].values\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, labels, test_size=0.2, random_state=42)\n",
    "\n",
    "label_binarizer = LabelBinarizer()\n",
    "y_train_cat = label_binarizer.fit_transform(y_train)\n",
    "y_test_cat = label_binarizer.transform(y_test)\n",
    "\n",
    "num_classes = 5  \n",
    "\n",
    "rnn_model = tf.keras.Sequential([\n",
    "    tf.keras.layers.Embedding(input_dim=20000, output_dim=32, input_shape=(X_train.shape[1],)),\n",
    "    tf.keras.layers.Bidirectional(tf.keras.layers.LSTM(128, return_sequences=True)),\n",
    "    tf.keras.layers.Bidirectional(tf.keras.layers.LSTM(128)),\n",
    "    tf.keras.layers.Dense(64, activation='relu'),\n",
    "    tf.keras.layers.Dense(num_classes, activation='softmax') \n",
    "])\n",
    "\n",
    "rnn_model.compile(loss='categorical_crossentropy',  \n",
    "                  optimizer=tf.keras.optimizers.Adam(1e-4),\n",
    "                  metrics=['accuracy'])\n",
    "\n",
    "# rnn_model.fit(X_train, y_train_cat, epochs=3,\n",
    "#               validation_data=(X_test, y_test_cat))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "bd3d7beb",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/5\n",
      "1500/1500 [==============================] - 441s 290ms/step - loss: 0.6950 - accuracy: 0.7139 - val_loss: 0.5072 - val_accuracy: 0.8110\n",
      "Epoch 2/5\n",
      "1500/1500 [==============================] - 433s 289ms/step - loss: 0.4082 - accuracy: 0.8536 - val_loss: 0.4350 - val_accuracy: 0.8481\n",
      "Epoch 3/5\n",
      "1500/1500 [==============================] - 433s 288ms/step - loss: 0.3193 - accuracy: 0.8931 - val_loss: 0.4130 - val_accuracy: 0.8580\n",
      "Epoch 4/5\n",
      "1500/1500 [==============================] - 436s 291ms/step - loss: 0.2636 - accuracy: 0.9141 - val_loss: 0.3950 - val_accuracy: 0.8673\n",
      "Epoch 5/5\n",
      "1500/1500 [==============================] - 438s 292ms/step - loss: 0.2209 - accuracy: 0.9303 - val_loss: 0.4281 - val_accuracy: 0.8665\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1200x500 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "history = rnn_model.fit(X_train, y_train_cat, epochs=5,\n",
    "                        validation_data=(X_test, y_test_cat))\n",
    "\n",
    "# Extract training history\n",
    "train_loss = history.history['loss']\n",
    "val_loss = history.history['val_loss']\n",
    "train_acc = history.history['accuracy']\n",
    "val_acc = history.history['val_accuracy']\n",
    "epochs = range(1, len(train_loss) + 1)\n",
    "\n",
    "# Plot training and validation loss\n",
    "plt.figure(figsize=(12, 5))\n",
    "plt.subplot(1, 2, 1)\n",
    "plt.plot(epochs, train_loss, 'g', label='Training Loss')\n",
    "plt.plot(epochs, val_loss, 'b', label='Validation Loss')\n",
    "plt.title('Training and Validation Loss')\n",
    "plt.xlabel('Epochs')\n",
    "plt.ylabel('Loss')\n",
    "plt.legend()\n",
    "\n",
    "# Plot training and validation accuracy\n",
    "plt.subplot(1, 2, 2)\n",
    "plt.plot(epochs, train_acc, 'g', label='Training Accuracy')\n",
    "plt.plot(epochs, val_acc, 'b', label='Validation Accuracy')\n",
    "plt.title('Training and Validation Accuracy')\n",
    "plt.xlabel('Epochs')\n",
    "plt.ylabel('Accuracy')\n",
    "plt.legend()\n",
    "\n",
    "plt.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "388decc1",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "id": "d66b2860",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 800x600 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "\n",
    "data_dict = {\n",
    "    'NB CountVectorize': 0.88,\n",
    "    'DT CountVectorize': 0.84,\n",
    "    'SVM CountVectorize': 0.89,\n",
    "    'NB Tfidf': 0.82,\n",
    "    'DT Tfidf': 0.84,\n",
    "    'SVM Tfidf': 0.93,\n",
    "    'LSTM': 0.93\n",
    "}\n",
    "\n",
    "categories = list(data_dict.keys())\n",
    "values = list(data_dict.values())\n",
    "\n",
    "plt.figure(figsize=(8, 6))\n",
    "plt.bar(categories, values)\n",
    "plt.xlabel('Model')\n",
    "plt.ylabel('Accuracy')\n",
    "plt.xticks(rotation=90)\n",
    "\n",
    "# Label the bars with their corresponding accuracy values\n",
    "for i, value in enumerate(values):\n",
    "    plt.text(i, value + 0.01, str(value), ha='center', va='bottom', fontsize=10)\n",
    "\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "34c91e36",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.10.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}