小世界无标度网络中考虑回音室的舆情演化  

Opinion dynamics model considering echo chamber effect in small-world scale-free networks

作　　者：周沁悦 吴志彬 徐玖平 Zhou Qinyue;Wu Zhibin;Xu Jiuping(Business School,Sichuan University,Chengdu 610065,China)

机构地区：[1]四川大学商学院,四川成都610065

出　　处：《系统工程学报》2023年第6期753-764,共12页Journal of Systems Engineering

基　　金：国家自然科学基金资助项目(71971148);中央高校基本科研业务费专项资金资助项目(2023ZY-SX023,2023ZYSX020,SXYPY202334).

摘　　要：基于社会网络分析和确认偏见心理,提出了回音室度量方法,进而建立了考虑回音室效应的舆情演化模型,设计网络生成模型,在局域优先连接的可变集聚系数网络中引入节点的适应性.实验结果表明,提出的网络生成模型能够产生具有小世界特性的无标度网络,能较好地模拟现实世界网络:利用生成网络和实际网络进行舆情演化模拟,与Hegselmann-Krause模型对比,结果表明,检测个体陷入回音室的倾向能够消除群体意见极化对立现象,且提高网络的集聚系数有助于减轻稳定意见中的回音室现象.This paper proposes an echo chamber measure for opinion formation based on social network anal-ysis and confirmation bias.Then the measure is used to build the opinion dynamics model considering the echo chamber effect.The proposed growing network model introduces the node fitness into the network with local-world preferential attachment and tunable clustering.Experimental results show that the proposed growing network model can generate scale-free networks with small-world characteristics.The simulations of the pro-posed opinion dynamics model and the Hegselmann-Krause model are both conducted in the proposed network model and real network,The results indicate that detecting the tendency of agents to fall into echo chambers can eliminate the opinion polarization.Besides,the networks with higher clustering coefficients produce lighter echo chamber effect.

关 键 词：社会网络 舆情动力学 无标度网络 小世界网络 回音室效应 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]