Asymmetry Opinion Evolution Model Based on Dynamic Network Structure

Abstract

On social media platforms, users can not only unfollow others whose opinion excessively opposes their own, but they can also add new connections. To better reflect the evolution of opinions on social media, this paper proposes an opinion asymmetry evolution model based on a dynamic network structure, where the trusts between two individuals are not mutual and dynamic. First, the paper analyzes the general properties of the model. We prove that group opinion can converge to a steady state even if the connection is unidirectional. Second, we compare the evolution process of static and dynamic network structures. Computer simulation results show that a higher probability of new connections leads to less aggregation of group opinion, higher information entropy, lower HHI, and lower degrees of polarization.