FakeBuster: Deceptive Content Identification Using Deep Learning, NLP, and Digital Forensics.

FakeBuster is a unified AI-powered multi-modal deceptive content detection framework integrating Natural Language Processing (NLP), Computer Vision, and Digital Forensics. The system detects fake news, deceptive reviews, forged documents, and deepfake videos using machine learning and deep learning techniques including TF-IDF with Multinomial Naive Bayes, Word2Vec with SVM, CNN-based image forensics, and CNN-LSTM deepfake analysis. The framework is deployed using a Flask REST API backend with a ReactJS frontend for real-time analysis and explainable AI-based output visualization. Experimental evaluation demonstrates strong performance across multiple deceptive content domains, with an overall average detection accuracy exceeding 90%. The project was developed as a final-year B.Tech research project at ADCET, Ashta.