Predicting, Evaluating, and Explaining Top Misinformation Spreaders via Archetypal User Behavior

A small number of highly influential users drive most misinformation spread: The research identifies that not all social media users contribute equally to misinformation—a small group of "super-spreaders" with distinct behavioral patterns can amplify false narratives and cause significant societal harm, suggesting targeted interventions could be more effective than broad-based approaches.

Different types of misinformation spreaders require different identification strategies: The paper identifies three distinct user archetypes—amplifiers, super-spreaders, and coordinated accounts—each with unique behavioral patterns. Super-spreader traits consistently ranked highest among influential misinformation spreaders, but the most effective approach recognizes that individual users often exhibit multiple archetype characteristics simultaneously.

Temporal patterns matter for predicting misinformation spreaders: The research demonstrates that tracking how users behave over time is critical for accurately predicting who will spread misinformation, and importantly, the models can work with shorter observation windows, reducing the data collection burden needed for early identification and intervention.

Explainable AI helps make misinformation detection more transparent and trustworthy: By using explainable AI techniques that integrate multiple behavioral indicators, the research shows how automated systems can identify problematic users while remaining interpretable—allowing platforms and researchers to understand why specific users are flagged rather than relying on opaque algorithmic decisions.