社交网络上舆论形成模型研究

发布时间：2018-01-09 12:20

本文关键词：社交网络上舆论形成模型研究　出处：《河南科技大学》2017年硕士论文　论文类型：学位论文

【摘要】：随着网络信息技术的快速发展,微博、微信、Facebook、Twitter等在线社交网络已成为人们获取信息、发表观点、表达情绪的重要途径,给舆论的传播、演化与形成带来极大便利。研究社交网络中舆论信息的传播过程和演化机制,对科学进行舆论的监控和引导具有重要的意义。网络舆情的相关研究逐渐从社会传播学、心理学延伸到了信息科学、复杂性科学、理论物理学等交叉学科领域。本文从复杂网络和协同理论视角出发,分析了社交网络舆论演化与形成过程中存在的内外因素,对舆论动力学模型进行了深入研究,主要工作包括:第一,针对规则网络、随机网络和小世界网络无法反映真实社交网络拓扑特性的问题,构建了具有可变聚类系数和无标度性质的社交网络模型,以此作为舆情演化模型的网络载体。第二,针对舆情演化过程中人际关系网络的复杂性和个体邻域的异质性,建立了协同舆情演化模型。通过不同参数下的实验,分别讨论了序参量和网络结构特征对系统磁化率的影响。结果表明,序参量更能体现舆情系统的协同效应,比较符合网络舆情演化的实际情况。第三,针对舆论传播和演化过程中个体交互的广泛性和个体社会影响力的差异性,在Hegselmann-Krause模型的基础上建立了社交网络舆论形成模型。新模型通过引入个体间亲密度、人际相似性和交互强度等概念,对个体交互集合进行了扩展,并对影响力权重进行了合理量化,进而构建更切合实际的观点交互规则。通过一系列仿真实验,对模型进行了深入分析和讨论。结果表明:在不同信任阈值下,群体观点均能收敛到一致,形成舆论共识;且信任阈值越大,收敛时间越短。同时,扩大交互集合、提高人际相似性的作用强度会促进舆论共识的形成。此外,当无标度网络的聚类系数和平均度较高时,群体观点更容易产生趋同效应。
[Abstract]:With the rapid development of Internet information technology, online social networks such as Weibo, WeChat Facebook and Twitter have become an important way for people to get information, express their opinions and express their emotions. It brings great convenience to the dissemination, evolution and formation of public opinion. This paper studies the communication process and evolution mechanism of public opinion information in social networks. It is of great significance to monitor and guide public opinion to science. The related research of network public opinion gradually extends from social communication, psychology to information science and complexity science. This paper analyzes the internal and external factors in the evolution and formation of social network public opinion from the perspective of complex network and cooperative theory. The main work is as follows: first, aiming at the problem that rule network, random network and small world network can not reflect the topology characteristics of real social network. The social network model with variable clustering coefficient and scale-free property is constructed as the network carrier of the evolution model of public opinion. Second. Aiming at the complexity of interpersonal relationship network and the heterogeneity of individual neighborhood in the process of public opinion evolution, a collaborative public opinion evolution model is established. The effects of order parameters and network structure characteristics on the magnetic susceptibility of the system are discussed respectively. The results show that the order parameters can reflect the cooperative effect of the public opinion system more in line with the actual situation of the evolution of network public opinion. Third. Aiming at the universality of individual interaction and the difference of individual social influence in the process of public opinion dissemination and evolution. Based on the Hegselmann-Krause model, a public opinion formation model of social network is established. The new model introduces the concepts of personal affinity, interpersonal similarity and interaction intensity. The individual interaction set is extended, and the weight of influence is quantified reasonably, and a more realistic interaction rule is constructed. A series of simulation experiments are carried out. The model is analyzed and discussed in depth. The results show that under different trust thresholds, the group views can converge to the same, forming consensus of public opinion; The larger the trust threshold, the shorter the convergence time. At the same time, expanding the interaction set and enhancing the intensity of interpersonal similarity will promote the formation of consensus of public opinion. In addition, when the clustering coefficient and average of scale-free network are higher. Group views are more likely to produce convergence effects.
【学位授予单位】：河南科技大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：O157.5;G206

【参考文献】