基于个体社会属性的网络舆情演化模型研究

发布时间：2018-01-05 11:08

本文关键词：基于个体社会属性的网络舆情演化模型研究　出处：《大连理工大学》2016年博士论文　论文类型：学位论文

【摘要】：当前,随着Web2.0技术的迅速发展,在线社交网络逐渐普及,舆情也逐渐的进行着线上网络和线下网络的融合,而且网络舆情的区域性特点也越来越显著。网络舆情事件一旦蔓延引发群体性事件,将极大的影响社会的稳定,典型的如2011年由日本福岛地震而引发的国内“抢盐事件”。因此,舆情的演化分析与预测得到了国内外众多学者的关注,已经成为当前研究的热点。然而,以往的大多数研究中,仅仅对单层网络中的舆情演化进行研究,而实际上,舆情不但在现实社会通过口口相传传播,也在微博、微信、BBS论坛等互联网平台上进行传播和演化；而且,由于网民结构、社会结构等构成的差异,使不同的区域中也展现了不同的舆情特点。因此,研究区域中线上网络和线下网络的舆情演化机制,是十分重要的。为此,本文针对区域舆情演化问题,从区域网民的个体社会属性构成角度出发,对在线社交网络、以及线上线下双层耦合网络上的舆情传播特点、传播机制、影响因素、传播模型等内容进行了研究。论文的主要研究工作如下：(1)统计分析了网络舆情发生的区域特点和传播的区域特点,在此基础上,构建了网络舆情社会脆弱性评估模型,实证结果表明我国的网络舆情社会脆弱性具有明显的区域特点。为准确评价面向网络舆情的社会脆弱性,从社会经济环境、网络信息交互水平、政府应急能力三个维度构建了包含12个三级指标的评价指标体系,运用熵权法和G1赋权法的组合赋权法对指标进行赋权,并对我国省级区域的社会脆弱性进行了实例分析。面向网络舆情的社会脆弱性评估是理解网络舆情风险和提高网络舆情应对能力的关键步骤。实例评价结果有助于在宏观上为区域的网络舆情预警提供参考。(2)研究了微博网络上舆情演化模型的构建方法。针对微博网络中只有转发节点才能对邻居节点的信息传播和观点演化有影响的特点,结合社会脆弱性理论,建立了融合微博网络中信息传播和观点交互模式的区域舆情动力学模型,探讨了微博网络中的舆情演化特征。首先,针对网民花费在网络上的时间和精力有限的特点,考虑微博用户添加好友的认知成本、反认知成本以及人的互惠心理,构建了微博网络模型。然后,考虑微博网络中的信息传播和观点交互特点,结合传播学中的社会心理学现象——社会加强效应和沉默螺旋效应,分别构建了微博网络上满足这两类传播心理学效应规则的舆情传播模型。第三,考虑微博网络中节点在传播路径上具有多层影响力、以及节点之间社会属性的不同而形成的不同自信度,提出了考虑多层邻居节点影响的微博网络舆情演化模型。第四,设置媒体作为外部影响因素,结合有界信任规则并考虑节点信任界限异质性的影响,研究了媒体效应对微博网络中舆情演化的影响,仿真实验结果同现实实例中的统计数据相吻合。(3)研究了线上线下耦合网络上区域舆情演化模型的构建方法。针对舆情在线下的现实社会网络和在线社交网络中双层甚至多层同时传播的特点,构建了考虑个体社会属性的线上线下耦合网络模型,并重点研究了该耦合网络上的舆情动力学特征。首先,分别考虑了线下现实社会网络的特点和在线社交网络的特点,根据节点社会属性之间的相似性选择连边,并通过一对一耦合的方式,构建了线上线下双层耦合网络模型,分析了该双层耦合网络的网络结构特征,并进一步仿真分析了谣言在该双层耦合网络上的传播与在单层网络传播的不同。然后,在上一步构建的双层耦合网络的基础上,考虑线上网络与线下网络的促进传播机制、以及节点社会属性不同而形成的不同自信度,提出了考虑个体社会属性特点的双层耦合网络演化模型。综上,本文的研究成果拓展了微博网络和多层耦合网络上的舆情动力学理论与方法,研究将有助于认识区域中的舆情演化特性,所提出的舆情演化模型对舆情演化的认知和舆情的预测与控制都具有较强的理论和现实意义。
[Abstract]:At present, with the rapid development of Web2.0 technology, the online social network popularization, public opinion is gradually carried out the fusion line network and line network, and regional characteristics of network public opinion is becoming more and more significant. Once the spread of public opinion events lead to mass incidents, will greatly affect the social stability, such as in 2011 by the Japanese earthquake caused the domestic "grab salt events". Therefore, the analysis and prediction of the evolution of public opinion has been the concern of many scholars at home and abroad, has become a hot topic in current research. However, most of the previous studies, only for the public opinion in the network layer evolution are studied, in fact, not only in public opinion the social reality through word of mouth spread, but also in micro-blog, WeChat, BBS Forum on the Internet platform for dissemination and evolution; moreover, due to the structure of Internet users, differences in social structure, The different regions show different characteristics of public opinion. Therefore, the research on the network public opinion center line and the line network evolution mechanism is very important. Therefore, according to the regional public opinion evolution, from the perspective of individual social attribute of regional perspective of Internet users, online social network, influence public opinion dissemination characteristics double coupling on the network, as well as online and offline communication mechanism, communication model has been studied. The main research work of this paper are as follows: (1) statistical analysis of regional characteristics of the regional characteristic and the dissemination of network public opinion, on this basis, build a vulnerability assessment model of network public opinion, the empirical results show that China's network public opinion social vulnerability has obvious regional characteristics. In order to accurately evaluate the social vulnerability of Internet public opinion, from the social and economic environment, the level of network information interaction, The three dimensions of government emergency capability construction includes 12 three class indexes, using a combination of entropy and G1 weighing method for index weighting, and on China's provincial social vulnerability are analyzed. For the network public opinion social vulnerability assessment is a key step in understanding the network public opinion risk and improve the network response capacity of public opinion. The evaluation result is helpful in the macro network public opinion warning area for reference. (2) research on the methods of constructing public opinion evolution model of micro-blog network. The network node can only transmit micro-blog information dissemination and the view of the neighbor node evolution characteristics of influence and the combination of social vulnerability theory, the establishment of a regional public opinion dynamics model of information transmission and fusion view micro-blog network interactive mode, discusses the evolution of the micro-blog network public opinion According to the characteristics. First, users spend on the Internet of the limited time and energy characteristics, considering the cognitive cost of micro-blog users to add friends, mutual psychological cognitive cost and anti people, build micro-blog network model. Then, considering the dissemination of information and interactive features of micro-blog in the network view, combined with social psychology phenomenon in communication effect and social effect to strengthen the spiral of silence, were constructed to meet the micro-blog network opinion propagation model of these two types of communication psychology effect rules. Third, considering the micro-blog network with multiple influence in the propagation path, and different social attributes between nodes and the formation of the different degree of self-confidence, micro-blog proposed based on multilayer network public opinion the neighbor node affects the evolution model. Fourth, set the media as the external factors, combined with bounded trust rules and consider trust boundaries The effect of heterogeneity, the media effect on public opinion evolution of micro-blog network, the simulation results of statistical data with practical examples in the match. (3) of the line construction method of regional public opinion evolution model under the coupling network. According to the reality of double public opinion online social networks and online social networks even in multilayer and at the same time the characteristics of transmission based on the consideration of individual social attributes of online and offline coupled network model, and focuses on the dynamic characteristics of the coupling of the network public opinion. Firstly, considering characteristics of the line under the reality of social networks and online social networks, according to the similarity between the social attribute node selection. The edge, and through a pair of coupled way, construction of the line under the double coupled network model, analyzes the network structure characteristics of the double coupling network, and further simulation Analysis of the spread rumors in the double coupling on the network and different in single network communication. Then, based on the double coupled network construction step on, consider the network online and offline network to promote the dissemination mechanism, different from reliability and social attributes of different nodes and the formation, considering the double coupling the network evolution model of individual social characteristics. In conclusion, the results of this study extends the opinion dynamics theory and method of coupling network of micro-blog network and multilayer the study will contribute to the understanding of public opinion in the area of the proposed public opinion evolution characteristics, evolution prediction and control model for the evolution of public sentiment cognition and has a strong public opinion the theoretical and practical significance.

【学位授予单位】：大连理工大学
【学位级别】：博士
【学位授予年份】：2016
【分类号】：C913.4

【相似文献】