Research & Papers

Graph-based framework detects disinformation narratives across Telegram channels

Weak supervision + propagation graphs uncover coordinated amplification in Ukraine-Russia info war.

Deep Dive

A graph-based framework is proposed for detecting disinformation narratives in Telegram ecosystems. The approach combines weak supervision with propagation graph analysis, aggregating semantically related claims into narrative-level clusters and modeling their diffusion across interconnected channels. Results demonstrate that integrating textual signals with network structure provides a scalable method for detecting disinformation narratives and offers insights into how they propagate within large-scale messaging environments.

Key Points
  • Combines weak supervision with propagation graph analysis to detect disinformation narratives
  • Aggregates semantically related claims into narrative-level clusters and models cross-channel diffusion
  • Integrated textual and network signals enable scalable detection of coordinated amplification

Why It Matters

Offers a scalable tool to uncover coordinated disinformation campaigns in messaging platforms, critical for information warfare monitoring.

📬 Get the top 10 AI stories daily