基于多智能体模型的微博舆论形成研究

On Public Opinion Formation of Microblogging in the View of Multi-agent Simulation

为从复杂系统视角探究微博舆论形成机制,根据微博议题内参与人的亲近性特征对意见更新方程进行了修正,并据此设定了个体行动规则,结合微博内议题参与者形成的网络拓扑结构,运用Repast平台进行仿真模拟,结果模拟出由个体意见的无序系统到有序系统的涌现,即形成舆论。最后用实例验证了模型的有效性。结论表明,微博内的舆论具有明显的沉默的螺旋和极化特征,在微博舆论尚未形成时,对其进行引导成本最低。

To explore the law of public opinion formation of micro-blogging from the perspective of complex systems, agent＇s action rules are set up. Opinion-updating equation is amended in accordance with affinity of participants of topics in micro-blogging. Combined with the network topology for the micro-blogging topics participants, the process of opinion formation is simulated. Disordered individual opi- ions emerged out of the system of ordering turns out to be the result, that is, forming public opinion. Example is used to verify the validity of the model. R will make exploratory groundwork for further media dissemination and monitoring of public opinion.