AI-Powered Fact Verification: Analyzing Public Consensus Through Machine Learning

In the face of rising misinformation across social media platforms, there is a growing need for scalable and automated systems that can verify factual accuracy. This study introduces a machine learning framework that leverages a transformer-based model to evaluate how well social media posts align with factual statements. The framework is tested across various classification strategies, including sentiment analysis, consensus scoring, and both binary and multi-class categorization. Among these approaches, binary classification demonstrated the strongest performance in detecting factual consistency. While sentiment-based methods offered limited insight, they were not sufficient to determine truthfulness on their own. The results highlight the potential of interpretable, high-accuracy models in supporting large-scale efforts to counter misinformation.