Opinion Dynamics on Adaptive Networks An Evolutionary Game Theoretical Approach with Incomplete Information
An evolutionary game with incomplete information is proposed to model the opinion formation on adaptive networks. Based on the bounded confidence model of continuous opinion dynamics, we introduce a new strategy that players will break their relationship and establish a new connection if their opinions differ more than the tolerance range. We take into account the discount of this action, which is private information for agents, however, the statistic characteristic is widely known by every one. In virtue of game theory of incomplete information we analyze Bayesian Nash equilibrium by using Harsanyi transformation. Simulation results show the macroscopically ordering process is accelerated in an incomplete information system, where few and large clusters are formed in the steady state and relaxation time is reduced. It is dramatic that with incomplete information opinion clusters and topology clusters become independent of bounded confidence. Although individual discount will get overt under the effect of memory, the influence over the ordering process is inapparent. Moreover, the model could be helpful for understanding some social phenomena of Internet consensus formation, and we provide valuable insights into Internet interactions.
opinion dynamics evolutionary game theory incomplete information adaptive network
Fei Xiong Yun Liu
Key Laboratory of Communication & Information Systems (Beijing Jiaotong University),Beijing Municipa Institute of Network Consensus, Beijing Jiaotong University, Beijing
国际会议
2010 Cross-Strait Conference on Information Science and Technology(2010 海峡两岸信息科学与技术学术交流会)
秦皇岛
英文
115-118
2010-07-09（万方平台首次上网日期，不代表论文的发表时间）