Opinion-Climate-Based Hegselmann-Krause dynamics

Abstract

Opinion dynamics, which focuses on the opinion evolution process of a group of agents, is an important problem in Cyber-Physical-Social Systems (CPSS). Although the classic Hegselmann-Krause model and its variants have been intensively studied, the opinion climate, which is the perceived majority of opinions of a given social group at a given time, has not been considered as a crucial factor in the evolution of agents’ opinions. In this paper, inspired by the theory of the spiral of silence, we propose a novel opinion-climate-based social Hegselmann–Krause dynamics model, where the opinion climate is quantified by calculating the relative proportion of people who hold positive or negative opinions. The evolution patterns of agents’ opinions that depends on not only himself and his neighbors, but also the environment are then recognized. Extensive experiments on both artificially generated and real Cyber-Physical-Social datasets show that when the opinion climate is taken into consideration, agents will eventually reach a consensus on some specific events. Moreover, the smaller the influence weight of the opinion climate, the greater the time required to reach a consensus. The results can well explain the group polarization phenomena we observed nowadays in our society and provide an important idea for Cyber-Physical-Social Services.