ICTMCG/fake-news-detection

The repository organizes peer-reviewed research across multiple detection approaches: social context analysis using graph neural networks and tensor factorization, multi-modal fusion combining text/image semantics, and content-based methods analyzing emotional, stylistic, and discourse patterns. It spans fact-checking, explainability, and transfer learning techniques—ranging from early propagation-path RNNs to recent meta-learning frameworks for emergent events with limited labeled data. The collection includes both foundational survey papers and implementation code, targeting researchers building detection systems across heterogeneous information networks, variational autoencoders, and active learning frameworks.

249 stars. No commits in the last 6 months.