Social media platforms are no longer just virtual town squares; they are high-stakes battlegrounds where algorithms, viral narratives, and coordinated toxicity directly trigger real-world action. Uncovering these underlying mechanics was the focus of the COSMOS Research Center at the 18th International Conference on Advances in Social Networks Analysis and Mining in Rabat, Morocco, on August 24-27, 2026. ASONAM is a premier global venue at the intersection of information systems, behavioral science, and big data analytics.

ASONAM is recognized internationally as a premier global forum for network science, big data analytics, and computational social science. The conference provides an essential platform for addressing critical challenges across information systems, focusing on algorithmic manipulation, social mobilization, dynamic toxicity modeling, and platform governance during acute geopolitical events. Presenting a major body of work at this highly competitive venue highlights COSMOS’s ongoing commitment to advancing data-driven tools that safeguard public discourse and foster digital resilience.

Publishing at ASONAM 2026 directly advances COSMOS’s mission to analyze online influence, cognitive security, and digital platform resilience, supported in part by major federal research grants dedicated to national defense and information stability. Through these projects, COSMOS researchers examine real-world social movements, algorithmic recommendation traps, and cross-platform narrative trajectories across key international contexts, including the 2025 Indonesian protests, YouTube recommendation networks, global blogging ecosystems, and multi-platform communication channels.

The 13 accepted studies fall into three main research clusters that push the boundaries of computational social science. The first cluster focuses on AI-based algorithmic recommendation systems and digital platforms, introducing newly updated analytics infrastructure including VTracker 2.0, BlogTracker 2.0, and DataVerse. These tools use large language models (LLMs) and network science to detect content traps, map evolving YouTube recommendation pathways, and manage multi-platform data at scale. The second cluster explores real-world social mobilization and narrative diffusion, examining how symbolic structures, emotional coherence, and key focal groups drove online and offline activity during the 2025 Indonesian protests. The third cluster addresses online toxicity, covert influence, and dynamic modeling, using epidemic propagation theories, multiplex network algorithms, and temporal forecasting models to trace how toxic behavior and cross-platform campaigns spread over time.

By combining advanced AI, computational network science, and human behavioral analysis, these studies offer practical frameworks for identifying online threats before they scale. Together, this research reinforces COSMOS’s commitment to creating transparent, secure, and resilient digital spaces that preserve democratic integrity worldwide.