International Journal of Scientific Research and Engineering Development

This research presents a comprehensive graph-based machine learning framework for modeling and predicting misinformation spread on social networks. The proliferation of false information across digital platforms has emerged as a critical challenge affecting public opinion, democratic processes, health outcomes, and social cohesion. Traditional content-based detection methods often fail to capture the complex network dynamics and propagation patterns that characterize misinformation diffusion. The proposed system integrates Graph Neural Networks (GNN) for network structure analysis, Natural Language Processing (NLP) models for content credibility assessment, and temporal sequence analysis using Long Short-Term Memory (LSTM) networks to track information cascade patterns. The framework employs node embedding techniques to represent users and content within a unified feature space, enabling the identification of influential spreaders and vulnerable communities. By combining structural, textual, and behavioral features, the model achieves superior performance in early detection of misinformation cascades before widespread propagation occurs. Experiments conducted on real-world social media datasets including Twitter, Facebook, and Reddit demonstrate that the proposed ensemble approach significantly outperforms baseline methods in accuracy, precision, and early warning capabilities. The results provide actionable insights for platform moderators and policymakers to implement targeted intervention strategies. This work contributes to computational social science by offering a scalable, interpretable solution for combating information disorders in online environments.

Dr.V.Kavitha, M.Abdul Hameethu, S.Vinoth Kumar,"Misinformation Spread Modeling on Social Networks Using Graph-Based Machine Learning" International Journal of Scientific Research and Engineering Development, V9(1): Page(1169-1176) Jan-Feb 2026. ISSN: 2581-7175. www.ijsred.com. Published by Scientific and Academic Research Publishing.