📰 Fake News Detection System

NLP-powered misinformation classifier using TF-IDF features and Logistic Regression to distinguish fake from real news.

---

📋 Overview

A machine learning pipeline that combats misinformation by automatically classifying news articles as real or fake. The system combines article titles and body text, applies TF-IDF vectorization, and uses Logistic Regression for binary classification with high accuracy.

🏗️ Pipeline

News Article (Title + Body)
    ↓
Text Preprocessing (combine title & text, lowercase, clean)
    ↓
TF-IDF Vectorization (stop words removal, frequency filtering)
    ↓
Logistic Regression Classifier (max_iter=1000)
    ↓
Prediction: REAL / FAKE

✨ Features

• 🔍 Title + Body Fusion — Combines headline and article text for richer features
• 📊 TF-IDF with Smart Filtering — max_df=0.95, min_df=5 to filter noise
• 🎯 Stratified Train/Test Split — Ensures balanced class distribution
• 📈 Full Evaluation Suite — Accuracy, Classification Report, Confusion Matrix
• ⚡ Fast Inference — Classify new articles in milliseconds

🚀 Quick Start

git clone https://github.com/Izumi6/Fake-News-Detection-System.git
cd Fake-News-Detection-System
pip install pandas scikit-learn
python fake_news_detection.py

🛠️ Tech Stack

Python · Scikit-learn · TF-IDF · Logistic Regression · Pandas

👤 Author

Suyash Vakhariya — suyashvakhariya.com · LinkedIn · GitHub