Text Classification for News Articles .ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "11f08fa1-e3ec-49bd-b934-aeab8fc8ffb3",
   "metadata": {},
   "source": [
    "# AG News Classification\n",
    "\n",
    "## Overview\n",
    "\n",
    "This project involves classifying news articles using the AG News dataset. The dataset consists of news articles categorized into different classes. The project demonstrates a text classification pipeline using natural language processing (NLP) and machine learning techniques.\n",
    "\n",
    "## Dataset\n",
    "\n",
    "The dataset includes two CSV files:\n",
    "- **Training Data (`train.csv`)**: Contains news articles used for training the model.\n",
    "- **Test Data (`test.csv`)**: Contains news articles used for evaluating the model.\n",
    "\n",
    "### Columns:\n",
    "- **Class Index**: Numeric index representing the class of the news article.\n",
    "- **Title**: Title of the news article.\n",
    "- **Description**: Full text of the news article.\n",
    "\n",
    "## Objectives\n",
    "\n",
    "### Data Exploration and Preprocessing:\n",
    "1. Load the dataset using pandas.\n",
    "2. Preprocess the text by tokenizing, converting to lowercase, and removing stopwords.\n",
    "3. Encode the categorical labels into numeric values.\n",
    "\n",
    "### Model Building and Evaluation:\n",
    "1. Implement a text classification model using a pipeline that combines TF-IDF vectorization and Multinomial Naive Bayes classification.\n",
    "2. Train the model on the training dataset.\n",
    "3. Evaluate the model on the test dataset using accuracy and classification metrics.\n",
    "\n",
    "### Model Persistence:\n",
    "1. Save the trained model for future use.\n",
    "\n",
    "## Methods and Tools\n",
    "\n",
    "### Libraries:\n",
    "- **pandas**: For data manipulation.\n",
    "- **nltk**: For text preprocessing.\n",
    "- **scikit-learn**: For building and evaluating the classification model.\n",
    "- **joblib**: For saving the trained model.\n",
    "\n",
    "### Algorithms:\n",
    "- **TF-IDF Vectorization**: Converts text into numerical features.\n",
    "- **Multinomial Naive Bayes**: A probabilistic classifier used for text classification.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "8f0bcd97-a9f8-4c08-99b3-2ad1cfc67e90",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy: 0.8935526315789474\n",
      "Classification Report:\n",
      "               precision    recall  f1-score   support\n",
      "\n",
      "           1       0.91      0.89      0.90      1900\n",
      "           2       0.95      0.97      0.96      1900\n",
      "           3       0.86      0.84      0.85      1900\n",
      "           4       0.86      0.87      0.87      1900\n",
      "\n",
      "    accuracy                           0.89      7600\n",
      "   macro avg       0.89      0.89      0.89      7600\n",
      "weighted avg       0.89      0.89      0.89      7600\n",
      "\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "['news_classifier_model.pkl']"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import pandas as pd\n",
    "from sklearn.feature_extraction.text import TfidfVectorizer\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.pipeline import Pipeline\n",
    "from sklearn.naive_bayes import MultinomialNB\n",
    "from sklearn.preprocessing import LabelEncoder\n",
    "from sklearn.metrics import classification_report, accuracy_score\n",
    "import joblib\n",
    "\n",
    "# Load data\n",
    "train_df = pd.read_csv('train.csv')\n",
    "test_df = pd.read_csv('test.csv')\n",
    "\n",
    "# Preprocessing\n",
    "def preprocess_text(text):\n",
    "    # Add your text preprocessing steps here\n",
    "    return text\n",
    "\n",
    "# Apply preprocessing to the text columns\n",
    "train_df['Description'] = train_df['Description'].apply(preprocess_text)\n",
    "test_df['Description'] = test_df['Description'].apply(preprocess_text)\n",
    "\n",
    "# Feature extraction and model pipeline\n",
    "pipeline = Pipeline([\n",
    "    ('tfidf', TfidfVectorizer()),\n",
    "    ('clf', MultinomialNB())\n",
    "])\n",
    "\n",
    "# Encode the labels\n",
    "label_encoder = LabelEncoder()\n",
    "y_train = label_encoder.fit_transform(train_df['Class Index'])\n",
    "y_test = label_encoder.transform(test_df['Class Index'])\n",
    "\n",
    "# Fit the model\n",
    "pipeline.fit(train_df['Description'], y_train)\n",
    "\n",
    "# Predict\n",
    "y_pred = pipeline.predict(test_df['Description'])\n",
    "\n",
    "# Calculate accuracy\n",
    "accuracy = accuracy_score(y_test, y_pred)\n",
    "print(\"Accuracy:\", accuracy)\n",
    "\n",
    "# Convert class_names to a list of strings if necessary\n",
    "class_names = [str(name) for name in label_encoder.classes_]\n",
    "\n",
    "# Print classification report with class names\n",
    "print(\"Classification Report:\\n\", classification_report(y_test, y_pred, target_names=class_names))\n",
    "\n",
    "# Save the model\n",
    "joblib.dump(pipeline, 'news_classifier_model.pkl')\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "2197b4d5-27ff-4118-85af-b94d7c131c25",
   "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": "3.11.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}