Temporal patterns

Analysis of temporal dynamics and patterns in the spread and engagement with fake news.

Key papers and articles

• Shu et al. (2019) — Hierarchical Propagation Networks for Fake News Detection: Characterizes temporal dynamics at multiple granularities: macro-level temporal features (time to influential user, lifespan, cascade origin intervals) show fake news has shorter lifespan and reaches amplifiers faster; micro-level temporal features (conversation lifespan, reply frequency) show fake news conversations are shorter; identifies temporal features as most discriminative for detection (F1 0.821 on PolitiFact macro, 0.826 on GossipCop).
• Ruchansky, Seo, & Liu (2017) — CSI: A Hybrid Deep Model for Fake News Detection: Uses temporal engagement frequency and inter-engagement time to capture response patterns; suspicious users identified as early promoters with burst engagement behavior.