基于多主体的微博舆情演化研究

发布时间：2018-03-09 09:32

本文选题：微博舆情　切入点：微博舆情演化　出处：《首都经济贸易大学》2017年硕士论文　论文类型：学位论文

【摘要】：随着网络社会的崛起以及微博用户量的急速增加,微博舆情逐渐成为一种不可忽视的频发性社会现象。通过聚合民意,微博舆情能够形成强大的社会合力,并产生深刻的社会影响。在此背景下,了解微博舆情的生成条件与环境,研究微博舆情的发展规律及影响因素,具有重要的现实意义。本文在梳理国内外相关研究成果的基础上,综合运用文献综述法、跨学科研究与计算机仿真等研究方法,分析一般微博舆情的演化过程。微博舆情演化是一个多主体协同的过程,具有小世界性属性,微博社区中相互关联的多样化行为主体通过关注、转发推送消息以自组织协同的方式推动着微博舆情的发展演化。本文通过梳理微博舆情、传染病动力系统模型、多主体仿真等方法理论,利用多主体建模对微博舆情的演化过程进行量化分析以及多主体仿真来对其演化过程进行一定程度上的模拟。本文的主要研究内容是:(1)阐述了微博舆情的内涵及其构成要素,并对其传播路径进行了说明,为接下来的微博舆情行为主体特性分析打下基础。(2)通过对微博舆情演化内涵的理解,引入自组织理论对微博舆情的演化动态过程进行刻画。(3)考虑微博舆情中行为主体的异质性,分析各主体之间的观点交互以及其相互影响因素和权重,利用多主体方法设定主体之间的交互规则,对传染病SIR模型中的概率进行更适用于微博舆情演化的改进。(4)利用多主体仿真软件Netlogo对微博舆情演化过程中的主体交互进行仿真,从主体交互及其相互影响的结果反映微博舆情演化的动态过程。(5)选取2015年“8·12天津港特大火灾爆炸事故”作为样本案例,通过数据的搜集整理,验证多主体仿真结果的正确性与客观性。通过本文对微博舆情演化动态过程的分析以及利用多主体仿真软件对其模拟刻画,找出微博舆情在其演化过程中各行为主体应注意的问题以及对政府相关部门更好地监测引导微博舆情提供一定的建议。
[Abstract]:With the rise of the network society and the rapid increase of Weibo's number of users, Weibo's public opinion has gradually become a frequent social phenomenon that can not be ignored. By aggregating public opinion, the public opinion of Weibo can form a strong social resultant force. And have a profound social impact. In this context, understand the conditions and environment of Weibo's public opinion, study the law of the development of public opinion and its influencing factors, It is of great practical significance. On the basis of combing the related research results at home and abroad, this paper synthetically applies the literature review method, interdisciplinary research and computer simulation research methods. Analysis of the evolution process of general Weibo public opinion. The evolution of public opinion is a multi-agent collaborative process, with a small global attribute. The diverse and interrelated actors in the Weibo community pay close attention to it. Forwarding and pushing messages promote the development and evolution of Weibo's public opinion in a self-organizing and cooperative way. This paper combs the theory of Weibo's public opinion, the model of infectious disease power system, the multi-agent simulation, etc. The evolution process of Weibo's public opinion is analyzed quantitatively by multi-agent modeling and the multi-agent simulation is used to simulate the evolution process to a certain extent. The main research content of this paper is: (1) expatiate the connotation and the constituent elements of Weibo's public opinion. And the transmission path is explained, which lays a foundation for the analysis of the main characteristics of Weibo's behavior of public opinion.) through the understanding of the connotation of the evolvement of public opinion of Weibo, The self-organization theory is introduced to depict the dynamic process of Weibo's public opinion evolution. (3) considering the heterogeneity of the behavior as the main body in Weibo's public opinion, this paper analyzes the interaction of viewpoints among the different subjects and their mutual influence factors and weights. The interaction rules between agents are set by multi-agent method, and the probability in SIR model of infectious diseases is more suitable for the improvement of Weibo's public opinion evolution. The simulation software Netlogo is used to simulate the agent interaction in the process of public opinion evolution. According to the dynamic process of Weibo's public opinion evolution reflected by the result of subject interaction and its interaction, this paper selects "August 12 Tianjin Port fire and explosion accident" on 2015 as a sample case, and collects and collates the data. Verify the correctness and objectivity of multi-agent simulation results. Through the analysis of Weibo public opinion evolution dynamic process and the use of multi-agent simulation software to describe its simulation, This paper finds out the problems that should be paid attention to by various actors in the evolution of Weibo's public opinion and provides some suggestions for government departments to better monitor and guide Weibo's public opinion.
【学位授予单位】：首都经济贸易大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：G206;C912.63

【参考文献】