Surprise Virality: Mapping and Predicting Unexpected Political Message Outbreaks on Social Media

While much of the literature (e.g., Lin & Tsai, 2022) emphasizes the role of algorithms and echo chambers in reinforcing polarization and filtering dissent, outlier events—where contrarian or cross-cutting political messages suddenly go viral—are underexplored. This project would use anomaly detection techniques (see Bennard et al., 2024) and large-scale social media data to systematically identify cases where political messages buck algorithmic expectations and gain unexpected traction. The research would then analyze content features, network structures, and temporal patterns to reveal the social or technological triggers behind these anomalies. By focusing on deviations from the expected, this study could uncover new mechanisms of agenda-setting and collective attention, challenging deterministic views of algorithmic control. The insights could help platforms and policymakers better understand—and perhaps foster—productive disruptions to polarization.

References:

1. Taking stock of social-political polarization in Asia: political communication, social media and digital governance. T. Lin, Chia-hung Tsai (2022). Asian Journal of Communication.
2. Cybersecurity on social media platforms: The use of deep learning methodologies for detecting anomalous user behavior. Adeoluwa Bennard, A. Babatope, Vishal Kumar, Seshagirirao Anil (2024). GSC Advanced Research and Reviews.