公共卫生突发事件中社会舆论形成、演化、引导、干预的系统建模与仿真分析

Regardless of whether belonging to the category of public health or public safety, one common ground can be summarized from the recent unconventional emergencies. That is the spread and dissemination of negative public opinion, and its effect on adding fuel to the fire and impact of deluding or confusing people. Therefore, system modeling and simulation analysis of public opinion formation, evolution, guidance, and intervention is necessary and urgent. The project focuses on the mechanism of public opinion propagation based on complexity theory and multi-agent simulation technology. Through embedding to the main integration platform and the 1st branch integration platform (simulation platform), serving for the application system will be provided for the master project. Results can be integrated through project researching are: 1) agent model in public opinion propagation; 2) information description model in public opinion propagation; 3) simulation model of public opinion propagation; 4) statistical individual attribute data, network opinion events data, and so on, which support parameters setting of public opinion model. The ultimate goal to achieve is bringing early warning and emergency management functions of public opinion into play, by the way of combining theory and practice, methods and technology.

反思和总结近年来的非常规突发事件，无论其性质是属于公共卫生类还是公共安全类，总能找出其中的一个共同点，这就是负面舆论的散布和传播，及其所带来的推波助澜效应和流言惑众影响。因此，社会舆论形成、演化、引导、干预的系统建模与仿真分析十分必要和紧迫。本项目基于复杂性理论和多主体仿真技术，梳理舆论传播的机理，为大项目中的应用系统类（模型）服务，落地到总集成平台+分集成平台一（仿真平台）里。期望通过项目研究可被集成的成果有：1）舆论传播中的Agent模型；2）舆论传播中的信息描述模型；3）舆论传播仿真模型；4）提供统计性个体属性数据、网络舆论事件数据等能够支持舆论模型的参数设置。最终目的旨在达到理论与实践、方法与技术相结合，发挥舆论的社会预警与应急管理等功能。

项目围绕“大项目”与“分集成平台一”的被集成需求与既定的研究目标进行了为期一年的探索和研究，全面完成了研究计划。研究过程中与集成项目充分交流，研究成果顺利对接到集成平台中。.本项目重点完成仿真集成平台对“非常规突发事件的信息处理与演化规律建模”的集成需求，探索完善了舆论仿真研究中两个实际问题，具体要点如下：.（一）针对舆论通用仿真模型的反演误差，依据实际案例（“新疆7•5暴恐”事件），定制Agent属性参数，通过梳理舆论事件特征抽取仿真参数，设计暴恐舆论扩散的Agent交互规则与仿真输出；.（二）针对舆论传播网络模型的单一化，构建“线上-线下”多层多属性耦合同步传播网络模型；并针对舆论管控策略研究与应用的鸿沟，完成舆论干预策略与心理疏导策略下仿真分析，以“杭州公交车纵火案”为例，对比不同情景下的社会效果。