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

The proliferation of misinformation on digital platforms necessitates scalable, automated truth-verification systems. This paper presents a machine learning-based framework that analyzes the alignment between social media posts and declarative statements using a transformer-based architecture. By experimenting with different classification paradigms—including consensus estimation, sentiment evaluation, and binary/multiclass predictions—the system’s capabilities and limitations are systematically assessed. Empirical results reveal that binary classification yields the highest accuracy, while sentiment-based alignment, though partially effective, fails to capture factual correspondence reliably. These findings contribute to designing robust, interpretable tools for large-scale misinformation detection.