Opinion dynamics on higher-order networks with stubbornness and trust

This paper proposes a novel opinion dynamics model based on two key psychological factors, namely, stubbornness and trust, that govern how agents update their opinions. By comparing the evolution of multiple configurations on hypergraphs, which capture group-based, higher-order interactions instead of pairwise ones, we find that heterogeneity leads to opinion fragmentation, whereas homogeneity drives the system toward consensus. This finding offers a plausible explanation for the persistence of opinion diversity in social networks. Through an analysis of opinion exchange between two opposing communities, we identify a group reinforcement effect driven by internal consistency, which effectively steers the direction of opinion flow. However, this reinforcement effect breaks down abruptly when a cluster's initial opinion strength falls below a critical point. This phase transition implies that achieving a critical opinion strength is a necessary condition for a weaker community to dominate a stronger one.