Final-Ensemble.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Importing Libraries"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy of PAC : 96.55%\n",
      "Confusion Matrix: [[2018   59]\n",
      " [  67 1513]]\n",
      "Multinomial Naive Bayes Accuracy: 77.96%\n",
      "Confusion Matrix: [[2073    4]\n",
      " [ 802  778]]\n",
      "Random Forest Accuracy: 85.89%\n",
      "Confusion Matrix: [[1970  107]\n",
      " [ 409 1171]]\n",
      "Ensemble Learning Accuracy: 89.99%\n",
      "Enter news content: This is shitttttt@@!!!\n",
      "Fake news\n"
     ]
    }
   ],
   "source": [
    "from sklearn.ensemble import VotingClassifier\n",
    "from sklearn.feature_extraction.text import TfidfVectorizer\n",
    "from sklearn.linear_model import PassiveAggressiveClassifier\n",
    "from sklearn.naive_bayes import MultinomialNB\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.metrics import accuracy_score\n",
    "import pickle\n",
    "import pandas as pd\n",
    "from sklearn.metrics import accuracy_score, confusion_matrix\n",
    "\n",
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "\n",
    "def load_dataset():\n",
    "    \"\"\"\n",
    "    Loads the fake news dataset and splits it into training and testing sets.\n",
    "    Returns the training and testing sets.\n",
    "    \"\"\"\n",
    "    # Load the dataset\n",
    "    df = pd.read_csv('train.csv').dropna()\n",
    "\n",
    "    # Split the dataset into training and testing sets\n",
    "    x_train, x_test, y_train, y_test = train_test_split(df['text'], df['label'], test_size=0.2, random_state=42)\n",
    "\n",
    "    return x_train, x_test, y_train, y_test\n",
    "\n",
    "\n",
    "\n",
    "def train_ensemble_model(x_train, x_test, y_train, y_test):\n",
    "    \"\"\"\n",
    "    Trains an ensemble model and saves it to a file.\n",
    "    Returns the trained ensemble model.\n",
    "    \"\"\"\n",
    "    # Define the models to include in the ensemble\n",
    "    pac = PassiveAggressiveClassifier(max_iter=1000)\n",
    "    mult_nb = MultinomialNB()\n",
    "    rfc = RandomForestClassifier(n_estimators=10, random_state=7)\n",
    "\n",
    "    # Train the models\n",
    "    tfidf_vectorizer = TfidfVectorizer(stop_words='english', max_df=0.7)\n",
    "    tfidf_train = tfidf_vectorizer.fit_transform(x_train)\n",
    "    tfidf_test = tfidf_vectorizer.transform(x_test)\n",
    "    \n",
    "    pac.fit(tfidf_train, y_train)\n",
    "    y_pred=pac.predict(tfidf_test)\n",
    "    score=accuracy_score(y_test,y_pred)\n",
    "    print(f'Accuracy of PAC : {round(score*100,2)}%')\n",
    "    print(\"Confusion Matrix:\",confusion_matrix(y_test,y_pred))#, labels=['0','1'])\n",
    "    \n",
    "    mult_nb.fit(tfidf_train, y_train)\n",
    "    MultNB_pred=mult_nb.predict(tfidf_test)\n",
    "    score1=accuracy_score(y_test,MultNB_pred)\n",
    "    print(f'Multinomial Naive Bayes Accuracy: {round(score1*100,2)}%')\n",
    "    print(\"Confusion Matrix:\",confusion_matrix(y_test,MultNB_pred))#, labels=['FAKE','REAL'])\n",
    "    \n",
    "    \n",
    "    rfc.fit(tfidf_train, y_train)\n",
    "    rfc_pred = rfc.predict(tfidf_test)\n",
    "    score2=accuracy_score(y_test,rfc_pred)\n",
    "    print(f'Random Forest Accuracy: {round(score2*100,2)}%')\n",
    "    print(\"Confusion Matrix:\",confusion_matrix(y_test,rfc_pred))#, labels=['FAKE','REAL'])\n",
    "    \n",
    "\n",
    "    # Create the ensemble and save it to a file\n",
    "    ensemble = VotingClassifier(estimators=[('PAC', pac), ('MultNB', mult_nb), ('RFC', rfc)], voting='hard')\n",
    "    ensemble.fit(tfidf_train, y_train)\n",
    "    filename = 'finalized_ensemble_model.sav'\n",
    "    pickle.dump(ensemble, open(filename, 'wb'))\n",
    "\n",
    "    # Compute the accuracy of the ensemble model\n",
    "    ensemble_pred = ensemble.predict(tfidf_test)\n",
    "    ensemble_score = accuracy_score(y_test, ensemble_pred)\n",
    "    print(f'Ensemble Learning Accuracy: {round(ensemble_score * 100, 2)}%')\n",
    "\n",
    "    return ensemble, tfidf_vectorizer\n",
    "\n",
    "\n",
    "def main():\n",
    "    # Load the dataset\n",
    "    data = pd.read_csv('train.csv').dropna()\n",
    "    X=data.text\n",
    "    Y=data.label\n",
    "\n",
    "    x_train, x_test, y_train, y_test = train_test_split(X,Y,test_size=0.2,random_state=7)# load_dataset()\n",
    "\n",
    "    # Try to load the trained ensemble model from file\n",
    "    try:\n",
    "        filename = 'finalized_ensemble_model.sav'\n",
    "        loaded_model = pickle.load(open(filename, 'rb'))\n",
    "        tfidf_vectorizer = pickle.load(open('tfidf_vectorizer.sav', 'rb'))\n",
    "    except FileNotFoundError:\n",
    "        # If the file doesn't exist, train a new ensemble model and save it to a file\n",
    "        loaded_model, tfidf_vectorizer = train_ensemble_model(x_train, x_test, y_train, y_test)\n",
    "        pickle.dump(tfidf_vectorizer, open('tfidf_vectorizer.sav', 'wb'))\n",
    "\n",
    "    # Use the trained ensemble model to predict the label of a new sample\n",
    "    text = str(input(\"Enter news content: \"))\n",
    "    text_vectorized = tfidf_vectorizer.transform([text])\n",
    "    prediction = loaded_model.predict(text_vectorized)\n",
    "\n",
    "    if prediction == 0:\n",
    "        print(\"True news\")\n",
    "    else:\n",
    "        print(\"Fake news\")\n",
    "\n",
    "\n",
    "if __name__ == '__main__':\n",
    "    main()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "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": "3.9.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}