外生变量和非正社会影响推动群体观点极化  

Exogenous covariate and non-positive social influence promote group polarization

作　　者：李振鹏[1,2] 唐锡晋[2] 

机构地区：[1]大理学院数学与计算机学院统计系,大理671003 [2]中国科学院数学与系统科学研究院系统科学研究所,北京100190

出　　处：《管理科学学报》2013年第3期73-81,共9页Journal of Management Sciences in China

基　　金：国家重点基础研究发展计划资助项目(2010CB731405);国家自然科学基金资助项目(71171187)

摘　　要：观点动向是群体行为研究的重要方面.群体的整体观点在个体间互动并在系统外因素共同影响下演化为不同模式.依据社会认同理论,给出了具体的3种社会影响,并在Hopfield网络模型中引入了三元组关系结构.模型仿真证实了群体外生变量的干涉和负社会影响是推动群体观点极化的主要因素,群体观点极化是社会影响和结构平衡共同作用的结果.从结构平衡理论角度出发揭示了网络全局稳定极化模式与局部结构平衡的关系.Opinion dynamics is one of the important aspects of crowd＇ s behaviors. Collective opinions evolve into different patterns, consensus, polarization or diversity under the cross-impact of a variety of factors, such as mutual influence, external intervention, etc. In this paper, three types of social influences are discussed from the perspective of social identity, and then we add triadic structure into Hopfield network model. Simula- tion results verify that exogenous covariate and non-positive social influence are the main factors to group polarization. Furthermore, we show that polarization is coexisting with local structure balance, and discover the intrinsic connection between global network stable pattern and local structure balance.

关 键 词：社会影响 观点极化 HOPFIELD网络 社会认同 结构平衡 

分 类 号：C936[经济管理—管理学] N945[自然科学总论—系统科学]