From Data to Insight: Hybrid Models for Fake News Detection using Generative AI and Web Comparison

The spread of false information in the digital age is a major concern. Traditional machine learning approaches are useful, but they often struggle to deal with the dynamic nature of incorrect information. This project aims to develop a hybrid fake news detection system that combines traditional machine learning, generative AI, and real-time online search comparison. Our approach begins with a logistic regression classifier that uses TF-IDF properties to quickly filter potentially incorrect content. The article’s linguistic and contextual coherence is then evaluated by an advanced analysis that makes use of OpenAI’s generative AI through the LangChain framework. A web search tool cross-references the content of the article with up-to-date online data to further assure accuracy. This multi-layered method increases detection accuracy and provides a comprehensive analysis of news veracity. Our results show significant improvements over existing methods and show the potential of combining web data, machine learning, and generative AI for precise fake news detection.