COSMOS continues to advance global scholarship in computational social science and network analytics with a comprehensive new survey published in the Journal of Social Network Analysis and Mining, a leading Springer Nature journal. This research examines the emerging field of “Computational Narratology“, which tracks how emotionally charged, structured stories propagate and evolve across online networks. Unlike traditional information diffusion models that treat social media posts as static, atomic units, this survey highlights the need for frameworks that capture the dynamic, interpretive nature of human storytelling. By synthesizing almost eight decades of interdisciplinary scholarship, the research introduces a detailed taxonomy of computational models, including narrative tracking models, role-based event chains, multimodal variational methods, and stance-aware epidemic adaptations, capable of tracking how complex storylines mutate, change in stance, and cross platform boundaries.

Key findings reveal that traditional information diffusion models fall short when capturing “semantic drift”, the phenomenon where stories continuously shift in tone, framing, and intent as they are remixed or satirized across polarized communities. The research demonstrates that modern narratives are inherently multimodal and multilingual, revealing that accurate tracking requires new cross-modal reasoning frameworks to follow storylines that blend text, memes, videos, and emojis while migrating across vastly different platform ecosystems.

The research highlights how narrative spread online is not merely the passive transmission of data but an active process where users act as co-authors, reshaping stories through collective interaction patterns and shared cultural resonance. By identifying critical gaps such as the lack of annotated narrative datasets, platform data restrictions, and the heavy use of sarcasm and figurative language, the survey establishes a methodological foundation for capturing the nuanced dynamics of online storytelling, offering a powerful tool for analyzing public discourse and identifying coordinated narrative manipulation.

This work underscores COSMOS’s leadership in AI-powered social media analytics, combining computational social science, network modeling, and ethical AI to solve pressing real-world problems. By publishing in Springer’s Journal of Social Network Analysis and Mining, COSMOS continues to contribute to international scholarship on narrative evolution, collective sensemaking, and the resilience of online ecosystems. Read the full article here.