A real-time fake news detection system built with an LSTM-based neural network and deployed via a Flask web app. It processes real-world news text input, predicts authenticity, and delivers results through a clean, responsive UI and RESTful APIs.
- π€ LSTM-based Neural Network built with Keras & TensorFlow
- π Achieved 92% accuracy on real-world news datasets
- π§ NLP preprocessing using NLTK and Keras tokenizers
- π Flask API backend with real-time prediction support
- π§ͺ OOP-based backend design with unit testing in Unix-based dev environment
- π Python
- π Keras, TensorFlow
- π NLTK, Pandas, NumPy
- π Flask (API)
- πΌοΈ HTML, CSS
- π§ͺ OOP, Unit Testing
- π» Unix-based CLI
- π Real-world news articles dataset (e.g., from Kaggle or trusted sources)
- Features include: title, text content, and label (fake or real)