虚拟社区网络中的极端个体建模与观点演化

摘要

In virtual communities, extremists often have a significant effect on the formation and development of group opinion. This paper introduces Receipt-Accept-Sample (RAS) model to community networks, builds an extremists model, and analyzes the formation and opinion dynamic of the agents'views. Simulation results show that the agents always have a trend to become extreme, while the probability of an extremist to change his view drops when his opinion awareness increases. The results also show that the proposed model fits the process of extremists' opinion dynamic.%在虚拟社区网络中,持极端观点的个体往往会对社区的群体观点的形成与发展产生更大的影响.本文将RAS模型拓展到社区网络上,建立极端个体模型,并分析其观点的形成与演化过程.分析与仿真结果表明,虚拟社区中的个体观点总是具有极端化的趋势,而一个极端个体发生观点转化的概率会随着观点系数的增加而快速减小.结果还说明,所提出的极端个体模型适合于描述网络社区环境下的极端个体的观点演化过程.