基于复杂网络的互联网舆情传递及监控研究

发布时间：2018-04-26 18:35

本文选题：舆情传递 + 阈值效应　；参考：《南京航空航天大学》2016年博士论文

【摘要】：互联网技术的发展和进步,极大的改变了人们的行为方式,尤其是在舆情的产生和传播上,其影响更是颠覆性的。近几年以来,由互联网舆情引起的突发群体事件此起彼伏,并随时有可能被激化,发展成为突发公共危机事件。由互联网舆情引起的群体事件频发,已经不是偶然现象,而是由个体利益诉求导致的必然现象。互联网信息高度的参与性、互动性、时效性以及病毒式传播的特点,使舆情的社会效应呈现几何的级数的放大,如果不对其进行适当干预,会经由互联网产生众多不良后果。然而目前对于网络舆情的监控仍然停留在被动管理的状态,尚未形成一种主动的事前预警和管理模式,是否可以“领先一步,预测需求”是有效引导和利用网络舆情的关键。本文在考虑互联网背景的前提下,通过对网络舆情传递理论的分析,希望寻找一种可行的网络舆情预测的方法,对网络舆情的管理者提供一些有益参考建议。本文在梳理国内外研究文献,概括网络舆情相关理论的基础上,分析了网络舆情传递的基本原理和路径,并对舆情控制方法和模式进行了探讨和研究,主要展开了以下6个方面的研究:⑴基于理论分析,对网络舆情的研究方法进行拓展,采用分层次分析的方式,将网络舆情整体层面、群体层面、个体层面三个层次分别进行建模研究和分析,建立从宏观到微观的全方位分析模型。⑵基于个体迁移和信息转换两个环节对互联网舆情传递过程进行理论解析,建立基于阈值的互联网舆情传递模型,同时仿真互联网舆情传播过程。⑶基于模型选择,通过互联网舆情传递过程中子群模式的解析,运用系统动力学的建模方法,来解释互联网舆情的传递过程,并发掘对互联网舆情子群传递模式产生影响的各种因素。⑷基于理论分析,对互联网舆情传递的总体模式做进一步的解析,从个体传递模式角度对互联网舆情传递过程来进行二次论述。运用系统动力学的分析范式,从个体间的互动视角对舆情个体迁移和信息转换两个环节进行模型建立和仿真。⑸基于复杂网络的互联网舆情传递模式分析,论述互联网舆情传递的事前监控问题。使用链路预测的分析方法,利用舆情传递网络的拓扑结构数据,以舆情个体间传递关系的“从众效应”和“阈值效应”为起点,来预测舆情个体间可能产生的舆情传递连接。⑹基于微博数据的互联网舆情现实测度,搜集经由微博传递的舆情现实数据,并对其进行网络整体测度、子群传递模式测度、个体传递模式测度,从而对本研究建立的舆情传递网络模型进行验证,从而证明其现实适用性。本文通过对网络舆情的系统研究,形成了一些新的研究结论和研究观点,主要的创新点集中在以下三个方面。⑴基于舆情传递理论分析,发现舆情传递过程的阈值效应,修正了传统的“一味压制”舆情控制观点。通过理论分析及舆情传递模型的建立,将舆情传递过程分解为舆情个体迁移及舆情信息转换两个阶段,并使用仿真数据对其进行模拟。仿真结果显示舆情传递过程即存在阈值空间,也存在阈值移动的小窗效应。改变传统的“一味压制”,依据阈值及阈值小窗效应来控制舆情传递更能实现舆情控制的目的,并且提高舆情个体的舆情观点转变概率是控制舆情传播的有效途径。从阈值角度提供了一个有效的舆情传播控制方法,以期为舆情监控提供相关的理论支持。⑵基于舆情传递过程分析,发现不同虚拟社区间及虚拟社区内部的互联网舆情传递模式的特点。本研究从虚拟社区角度出发,解析了互联网舆情传递的子群传递模式,着重研究互联网舆情在不同的虚拟社区间以及虚拟社区内部舆情传递的主要过程。⑶基于案例分析,对互联网舆情传递的“事前预测”做出尝试。目前,已有研究对互联网舆情的监控大多是事后角度的监控,通常的做法是对所收集的舆情数据进行统计分析或者传播动力学分析,找出其中的规律,及其可控之处。本研究尝试使用复杂网络理论中的“链路预测”方法,对互联网舆情传递过程进行事前预测,根据舆情网络的相关特征,选择适当的链路预测指标,完成舆情传递过程的整体预测。
[Abstract]:The development and progress of Internet technology have greatly changed people's way of behavior, especially in the production and dissemination of public opinion, its influence is more subversive. In recent years, the sudden group events caused by Internet public opinion have come and fall, and are likely to be intensified at any time to become a public crisis. The frequent occurrence of group events is not an accidental phenomenon, but an inevitable phenomenon caused by the appeal of individual interests. The highly participatory, interactive, timeliness and viral transmission characteristics of the Internet information make the social effect of the public opinion magnify the geometric series of the public opinion. If it is not properly intervened, it will be generated via the Internet. There are many adverse consequences. However, the monitoring of network public opinion remains in the state of passive management, and it has not yet formed an active pre warning and management model. It is the key to effectively guide and utilize the network public opinion on the basis of the internet background. The analysis of the theory of public opinion transmission, hope to find a feasible method of network public opinion prediction, and provide some useful references for the managers of network public opinion. On the basis of combing the research literature at home and abroad, summarizing the related theories of network public opinion, this paper analyzes the basic principles and paths of network public opinion transmission and the control methods of public opinion. And the model has been discussed and studied, and the following 6 aspects are mainly carried out: (1) based on theoretical analysis, the research methods of network public opinion are expanded, and the method of hierarchical analysis is adopted to study and analyze the overall level of network public opinion, group level and individual level by three levels, and to establish from macro to microcosmic. Based on the two links of individual migration and information conversion, the process of Internet public opinion transmission is analyzed theoretically, the threshold based Internet public opinion transmission model is established, and the communication process of Internet public opinion is simulated. (3) based on the model selection, the analysis of the model of neutrons in the process of Internet public opinion transmission is made, and the application system is applied. The modeling method of unified dynamics is used to explain the transmission process of Internet public opinion, and to explore the various factors that affect the transmission mode of the Internet public opinion subgroups. (4) based on the theoretical analysis, the overall model of the Internet public opinion transmission is further analyzed, and two comments on the transmission process of Internet public opinion are carried out from the perspective of individual transmission mode. In this paper, the model of system dynamics is used to model and simulate the two links of individual migration and information conversion from the interactive perspective between individuals. The network topology data, starting from the "herd effect" and "threshold effect", is used to predict the possible public opinion transfer connection between individuals in public opinion. Based on the actual measurement of Internet public opinion based on micro-blog data, it collects the real data transmitted by micro-blog through the public opinion and carries out the network as a whole. Measurement, subgroup transmission pattern measure, individual transfer model measure, thus verifying the public opinion transmission network model established in this study, thus proving its practical applicability. This paper has formed some new research conclusions and research views through the systematic research on network public opinion, and the main innovation points are concentrated in the following three aspects. (1) Through theoretical analysis of public opinion transmission, the threshold effect of public opinion transmission process is found, the traditional "one flavor suppression" public opinion control viewpoint is amended. Through the theoretical analysis and the establishment of public opinion transfer model, the process of public opinion transfer is divided into two stages of public opinion individual migration and public opinion information conversion, and simulation data is used to simulate it. The result shows that there is a threshold space in the process of public opinion transmission, and there is a small window effect of threshold movement. It is an effective way to control public opinion control to change the traditional "one flavor suppression", control the public opinion control according to the threshold and threshold small window effect, and to improve the opinion point change probability of public opinion. An effective method of public opinion communication control is provided from the threshold point of view to provide relevant theoretical support for public opinion monitoring. (2) based on the analysis of public opinion transmission process, the characteristics of Internet public opinion transmission in different virtual communities and virtual communities are found. Delivery of subgroup transfer mode, focusing on the main process of Internet public opinion in different virtual communities and virtual community internal public opinion transmission. 3. Based on case analysis, the "pre prediction" of the Internet public opinion is tried. At present, the research on Internet public opinion is mostly monitored after the event. This study tries to use the "link prediction" method in the complex network theory to predict the transmission process of Internet public opinion, and to select appropriate links according to the characteristics of public opinion network. Prediction indicators to complete the overall forecast of public opinion transfer process.

【学位授予单位】：南京航空航天大学
【学位级别】：博士
【学位授予年份】：2016
【分类号】：G206

【相似文献】