Purpose The widespread use of social media has increased the complexity of public opinion dissemination networks, making effective monitoring and regulation more challenging. Identifying key factors that trigger opinion outbreaks in a timely manner is crucial for crisis management. This study develops a dissemination model based on multi-tiered circle structures (MCDM) to analyse hierarchical public opinion diffusion on social media. Design/methodology/approach Using dissemination data from “the Big-Headed Baby incident involving antibacterial cream” public opinion event on Sina Weibo, collected between January 6 and February 6, 2021, this study employs complex social network analysis and time-series modelling to construct overlapping multi-tiered dissemination structures. Using core-periphery analysis, we propose a dynamic method to identify key users across different hierarchical tiers. Additionally, we examine the roles of sentiment evolution and cross-circle interactions in shaping public opinion dissemination. Findings The results indicate that (1) as hierarchical depth increases, dissemination volume grows exponentially; (2) in smaller circles, user interaction levels and cross-circle engagement are negatively correlated with propagation distance and (3) abrupt shifts in negative sentiment frequently occur within multi-tiered dissemination structures, significantly influencing information spread. Cross-circle interactions play a crucial role in opinion diffusion. Originality/value This study unveils the circle-based phenomenon in social media opinion dissemination through the dynamic identification of multi-tiered dissemination structures and cross-circle interactions, offering new perspectives for public opinion governance. It proposes the MCDM, which integrates sentiment, hierarchical depth and cross-circle interactions to explain dissemination dynamics.
Zhu et al. (Tue,) studied this question.