情绪驱动的网络群体行为演化机理与政府引导策略研究

Network information is often an emotional expression. On the Internet, an event from occurring, developping, spread to the crowd, is not only the process that network group emotion is excited, infected and exponential explosion, but also the emergence of individual behavior to group behavior. The project will analyze the influence mechanism of emotion to network individual behavior based on the theory of sociology and psychology, identify and concise network subject attributes according to collection and collation to a large amount of network group events information, explore the emotion dissemination and diffusion mechanism in the environment of network on basis of rumor spreading model and technology diffusion process theory. Based on multi agent computational experimental method, the project will build a computational experiment system that can describe the characteristics of network subject, so as to study the network group emotion communication and diffusion path in different network structure, and the influence of the emotion transmission mechanism to evolution of network group behavior. Then we can present scientific prediction and effective control of group emotional evolution, and put forward some pertinent government guidance strategies to mitigate the unfavorable effects that extreme negative emotions may bring. It enriches the application of computational experimental method in the social sciences especially the public management field which has both theoretical value and practical significance.

网络信息常是一种情绪的表达，在互联网上，一个事件从发生、发展、扩散到群体聚集，既是网络群体情绪被激发、感染和指数级爆发的过程，也是个体行为向群体行为涌现的过程。本项目将根据社会学和心理学的研究成果分析情绪对网络个体行为的影响机理、通过对大量网络群体事件信息进行收集和整理来识别和凝练网络主体的属性、根据谣言传播模型和技术扩散过程理论研究情绪在网络环境中的传播和扩散机理、基于多Agent的计算实验方法构建能反映网络主体特征的计算实验模型系统，从而分析不同网络结构下网络群体情绪的传播、扩散路径以及情绪传导机制对群体行为演化模式的影响，进而科学预测并有效控制群体情绪的演化结果，通过博弈模型与计算实验系统研究政府行为并提出有针对性的政府引导策略来及时消除极端的群体负面情绪可能带来的不利影响，这对于丰富计算实验方法在社会科学特别是公共管理中的应用既有理论价值又有现实意义。

项目以计算实验方法和实证分析方法为主，研究网络群体情绪的传播与演化规律，网络谣言传播中的话题扩散与演化规律，大型工程利益主体网络、环境问题中的公众认知网络、上市公司信息网络等情绪传播等问题。主要贡献有：（1）在网络群体行为演化机理研究中，引入计算实验方法对网络谣言传播问题进行研究，发现网络谣言传播过程中的内容决定媒介、遗忘机制、话题跳跃性传播、观点交互随群体一致性动态变化等规律。（2）运用事件分析法及回归分析并结合运用定性与定量的研究方法，探索上市公司负面网络舆情与公司价值之间的内在关联关系。（3）分析网民群体态度演变的机制，得到网民群体态度与初始认知相关，群体态度转变率与话题内容无关等结论。（4）引入空间系统动力学中的协同演进建模方法，通过仿真实验探索了系统中平台、媒介和网民主体对网络谣言传播的影响，得到网络谣言演化的一般规律。