双层耦合网络中基于改进SIHR模型的舆情传播研究

Research on Public Opinion Propagation in Coupled Double-Layer Network Based on Improved SIHR Model

摘要: 当前信息网络迅速发展，舆情在社交网络上的传播变得十分复杂。舆情传播的现实情况表现为多个社交网络间共同传播，单个社交网络的舆情传播研究已不能真实地描述现实情况。除此之外，舆情传播过程中，个体状态的变化存在多样性。特别的是，存在初次“免疫”舆情的个体由于后续产生兴趣而变为传播者的情况。因此，本文将以双层耦合网络为载体，基于改进的SIHR模型来研究舆情传播动力学过程并提出具体的应对策略。具体为，首先根据改进的SIHR模型，给出各状态人群在双层网络下的状态转移图。再根据离散的马尔可夫链方法，给出改进的SIHR模型的舆情传播动力学方程。最后，通过MATLAB软件开展模拟仿真，来研究舆情传播的规律及模型参数对于传播过程的影响。并在此基础上，从公共部门、社交平台、用户三个方面给出舆情应对策略，包括：强化官方媒体的威信力，加强平台的监管力度，做好用户情感引导。

Abstract: With the rapid development of information network, the diffusion of public opinion on social networks has become very complicated. The actual situation of public opinion diffusion is manifested as the common communication among multiple social networks, and the research on public opinion diffusion of a single social network can no longer truly describe the real situation. In addition, in the process of public opinion dissemination, there is diversity in the change of individual status. In particular, there are cases where individuals who are initially “immune” to public opinion become communicators due to subsequent interest. Therefore, this paper will study the dynamic process of public opinion diffusion based on the improved SIHR model and put forward specific countermeasures with the carrier of double-layer coupled network. Specifically, according to the improved SIHR model, the state transition diagram of each state population in the two-layer network is given. Based on the discrete Markov chain method, the public opinion propagation dynamics equation of the improved SIHR model is given. Finally, MATLAB software is used to carry out simulation to study the law of public opinion propagation and the influence of model parameters on the propagation process. Based on that, from the three aspects: the social platform, public sector and users, we give public opinion controlling strategies, including: strengthening the supervision of the platform, enhancing supervision of the official media and effectively guiding user emotions.