Skip to content

README.md

Latest commit

 

History

History
156 lines (97 loc) · 3.62 KB

README.md

File metadata and controls

156 lines (97 loc) · 3.62 KB

📰 Fake News Detection using LSTM (Deep Learning)

🚀 An end-to-end AI-powered Fake News Detection system built using LSTM Neural Networks and deployed as a live web application on Hugging Face. Fake_news_detector_thumbnail

📌 Project Overview

With the rapid growth of digital media, misinformation and fake news spread quickly across social platforms.
This project aims to build a deep learning-based system that classifies news articles as REAL or FAKE using Natural Language Processing (NLP).

The system is trained on labeled news data and deployed as an interactive web application.

🎯 Features

  • ✅ Text preprocessing and cleaning pipeline
  • ✅ LSTM-based Deep Learning model
  • ✅ ~95% validation accuracy
  • ✅ Real-time prediction via Streamlit Web App
  • ✅ Deployed on Hugging Face Spaces
  • ✅ End-to-end ML workflow implementation

🧠 Model Architecture

The deep learning architecture consists of:

  • Embedding Layer – Converts words into dense vectors
  • LSTM Layer (64 units) – Captures sequential dependencies
  • Dropout Layer (0.3) – Prevents overfitting
  • Dense Output Layer (Sigmoid) – Binary classification (Real/Fake)

Loss Function: Binary Crossentropy
Optimizer: Adam

🔄 Workflow

  1. Data Collection (Kaggle Fake News Dataset)
  2. Data Preprocessing
    • Remove special characters
    • Lowercasing
    • Tokenization
    • Padding sequences
  3. Model Training using TensorFlow/Keras
  4. Model Saving (.keras format)
  5. Streamlit App Development
  6. Deployment on Hugging Face Spaces

📊 Results

  • 📈 Training Accuracy: ~96%
  • 📈 Validation Accuracy: ~95%
  • ⚡ Real-time prediction performance
  • 🌍 Successfully deployed online

🌐 Live Demo

🔗 Hugging Face Deployment Link:

https://huggingface.co/spaces/RakeshBabuGajula/Fake-News-Detector Screenshot 2026-02-21 172234 Screenshot 2026-02-21 172943 Screenshot 2026-02-21 172800 Screenshot 2026-02-21 173009

🛠️ Tech Stack

  • Python
  • TensorFlow / Keras
  • Scikit-learn
  • Pandas & NumPy
  • Streamlit
  • Hugging Face Spaces

📁 Project Structure

fake-news-detector/ │ ├── app.py ├── fake_news_model.keras ├── tokenizer.pkl ├── requirements.txt ├── README.md

🚀 Installation (Local Setup)

Clone the repository:

git clone https://github.com/your-username/fake-news-detector.git

cd fake-news-detector

Install dependencies:

pip install -r requirements.txt

Run the application:

streamlit run app.py

🔮 Future Improvements

  • Upgrade to BERT / Transformer model
  • Add multilingual support
  • Add explainable AI (highlight suspicious words)
  • URL-based news prediction
  • Mobile app version

📚 References

  • TensorFlow Documentation
  • Streamlit Documentation
  • Hugging Face Spaces
  • Kaggle Fake News Dataset

👨‍💻 Author

Rakesh Babu Gajula
B.Tech - Computer Science Engineering
AI & Machine Learning Enthusiast

⭐ If you like this project, give it a star!