Global narrative events show recurring burst patterns across topics
Peak-DNA: Recurrent Morphology in Global Narrative Propagation
Cryptography and Security
Summary
Narrative news events around the world show bursts of activity that are noisy and irregular, making them hard to compare directly. The authors created Peak-DNA, a method that simplifies and aligns these bursts into clearer patterns and then compares them to detect repeated shapes of amplification. They tested this on stories about Iran, defense, AI, space, and quantum topics, finding that events within each topic tend to follow similar burst shapes over time. This work reveals a new way to track the structure of news spread beyond just what the news is about or how big a story is.
What this means in practice
- •For news monitoring teams: Identify repeated global news spread patterns within specific topics to improve event tracking and prediction accuracy.
- •For social media analysts: Detect structural similarities in how narratives amplify across platforms despite noisy and asynchronous data sources.
Authors
Joosung Kim, Luyonghe Li, Shanwen Wu
Abstract
Global narrative streams exhibit noisy, fragmented, and asynchronous amplification bursts, making direct waveform comparison unreliable for recurrence analysis. This paper introduces Peak-DNA, a framework for identifying recurrent propagation morphology in global narrative streams. The proposed framework transforms hourly GDELT activity into normalized macro-peak episodes through denoising, macro-peak consolidation, and temporal alignment, and compares propagation morphology using DTW-based similarity analysis. Using GDELT narrative streams, we analyze five high-amplification narrative domains: Iran, Defense, AI, Space/Elon, and Quantum/PQC. The results show that amplification events repeatedly converge toward morphology-level propagation patterns within each narrative domain. DTW-based similarity analysis further demonstrates strong intra-domain structural compactness despite heterogeneous publication dynamics and severe observational noise. These findings suggest a measurable morphology layer beyond semantic content or anomaly intensity, providing an initial structural foundation for recurrence-aware narrative monitoring, propagation family discovery, and longitudinal Narrative Genome analysis.