Scalable AI-Driven Analytics for User Engagement and Stance Detection on Social Media

Collection of recent arXiv papers (May 29 2026) on AI and social systems. Key contributions include: a scalable, service-oriented pipeline for large-scale user engagement and stance detection on social media (7M+ YouTube comments) exposing rapid early amplification of conspiracy content and concentrated high-activity users; a deep reinforcement learning approach for fairness-aware profit maximization in influence problems; SAHG, an anisotropic hyperbolic-geometry graph model for robust social-bot detection that separates account and neighborhood channels; empirical analysis of attention asymme