社会网络的信息传播机制及控制方法研究

发布时间：2018-04-24 02:07

本文选题：社会网络 + 信息传播　；参考：《哈尔滨理工大学》2016年博士论文

【摘要】：随着Web2.0理念和技术的日趋成熟,社会媒体也正逐渐被知晓、理解、熟悉、认可和应用。以Facebook、Twitter、新浪微博和微信等为代表的社会媒体,已经改变了人们传统的生活方式,更是成为彼此分享生活、状态、观点、经验和兴趣的平台。社会媒体的内在结构称为社会网络,社会网络以其用户数量巨大、数据类型广泛、传播速度迅猛等特征而著称。社会网络中的海量信息呈现交叉网状传播,传播速度和受众量呈指数级增长。因此,社会网络也随之点燃了工业界和学术界的研究热情。首先,能否对社会网络的信息传播数据进行观测与跟踪,并精确分析其传播规律,已经成为社会网络研究的重中之重;其次,是否存在信息传播模型可以精准刻画信息传播规律,也成为研究重点;同时,信息传播整体态势能否被感知?信息传播行为能否被预测?影响信息传播预测的相关因素有哪些?可否建立信息传播预测模型?这些都是开放性并极具挑战性的问题;此外,社会网络的信息传播是否可控?如何让信息控制的成本最小化,如何选择最佳控制点以便控制大部分甚至整个网络,都存在很大的研究空缺;最后,如何将各个方法完美融合并集成到实际应用系统中,如何研发出社会网络信息传播分析及控制系统,仍需进一步的综合验证。因此,基于上述研究背景及相关问题,本文的主要工作包括:1.在社会网络的信息传播规律研究方面阐述了信息传播机制和社会网络结构对信息传播的影响,提出了一种基于传播递归的社会网络信息传播树生成算法,基于对新浪微博真实案例的洞察和观测结果,归纳并总结了七种社会网络信息传播模型;此外,在充分分析信息传播过程与传染病传播过程的异同后,引入正向感染及负向感染两个状态,提出了一种新型社会网络的信息传播模型(SPNR)。2.在社会网络的信息传播预测研究方面首先介绍了社会网络的信息传播态势感知问题的基本思想,并建立了信息传播态势与Modularity指标的映射关系,提出了一种基于三次指数平滑的信息传播态势感知模型;此外,针对信息爆发预测问题,提出了一种信息爆发预测问题的定义方法,从拓扑结构变化、时序演变、用户属性和内容演化四个方面深度分析了影响信息传播预测的相关因素,并提出了信息爆发预测模型。3.在社会网络的信息传播控制研究方面首次证实了级联失效现象不仅仅出现在传统网络中,社会网络也同样存在级联失效现象,并且分析了级联失效现象与信息控制之间的关系;发现了社会网络中存在一些“超级用户”能够促进级联失效现象的演变,提出了一种基于信任危机的社会网络级联失效模型,该模型可以实现在只控制很少用户节点的情况下,使得用户之间产生信任危机并逐级失效,从而达到控制大部分甚至整个网络的效果。4.在社会网络信息传播分析及控制系统方面基于社会网络的信息传播规律、预测和控制三方面的研究成果,设计并实现了Woodpecker系统。详细阐述并分析了Woodpecker的系统架构和处理流程,并介绍了Woodpecker的核心模块,包括:信息数据获取模块、信息传播分析模块、信息传播预测模块以及信息传播控制模块。此外,通过对Woodpecker的功能和性能进行测试,发现Woodpecker能够对社会网络的信息传播规律进行有效地分析、预测以及控制。本文可以更好地理解社会网络的信息传播规律、有效地对社会网络的信息传播态势及爆发行为进行预测、精细化地对社会网络的信息传播进行控制。其研究成果在信息传播、商业营销、信息推荐、社会管理等多个领域都具有很强的参考价值及借鉴意义。
[Abstract]:With the growing maturity of Web2.0 concepts and technologies, social media is gradually known, understood, familiar, recognized and applied. Social media, represented by Facebook, Twitter, Sina micro-blog and WeChat, have changed the traditional way of life and become a platform for sharing life, state, views, experience and interest to each other. The internal structure of the body is known as the social network. The social network is famous for its huge number of users, extensive data types and rapid propagation speed. The mass information in the social network is intersected with an intersecting network, and the speed of communication and the audience are increasing exponentially. Therefore, the social network also ignites the research heat of industry and academia. First, whether the information dissemination data of the social network can be observed and tracked, and the accurate analysis of its propagation law has become the most important part of the social network research. Secondly, whether the information dissemination model can accurately depict the law of information dissemination and become the focus of the research; at the same time, can the overall situation of information spread be perceived? Can information dissemination behavior be predicted? What are the factors affecting the prediction of information dissemination? Can it be possible to establish an information dissemination prediction model? These are open and challenging issues. In addition, is the information dissemination of the social network controllable? How to minimize the cost of information control and how to select the best control points to control the large part In the end, how to integrate each method perfectly and integrate into the actual application system, how to develop the social network information dissemination analysis and control system is still needed further comprehensive verification. Therefore, based on the research background and related problems, the main work of this paper includes: 1. In the research of information dissemination law of social network, the influence of information dissemination mechanism and social network structure on information dissemination is expounded. A new algorithm for generating information propagation tree based on communication recursion is proposed. Based on the insight and observation results of sina micro-blog real cases, seven kinds of social network information transmission are summarized and summarized. In addition, after analyzing the similarities and differences between the information transmission process and the transmission process of infectious diseases, the two states of positive infection and negative infection are introduced, and a new social network information communication model (SPNR).2. is proposed in the research of the information dissemination of social network. The basic idea of the problem and the mapping relationship between information dissemination situation and Modularity index are established. A information dissemination situation perception model based on three times exponential smoothing is proposed. In addition, in view of the problem of information burst prediction, a new method for information outburst prediction is proposed, from topological structure change, time series evolution and user genera. In the four aspects of nature and content evolution, the related factors affecting the information dissemination prediction are analyzed, and the information burst prediction model.3. is proposed for the first time in the research of the information communication control of the social network. The cascade failure phenomenon is not only appeared in the traditional network, and the social network also has cascading failure phenomena, and the analysis of the network also has the same cascading failure phenomenon. The relationship between cascaded failure and information control is found. It is found that some "super users" can promote the evolution of cascading failures in social networks, and a cascaded failure model of social networks based on trust crisis is proposed. This model can make users produce only a small number of user nodes. In order to control the effect of most even the whole network,.4. has designed and implemented the Woodpecker system in the three aspects of the social network information communication analysis and control system, which is based on the social network information dissemination law, the prediction and control of the social network information communication analysis and control system. The system of Woodpecker is elaborated and analyzed in detail. The core module of Woodpecker is introduced, including information data acquisition module, information communication analysis module, information communication prediction module and information communication control module. In addition, through testing the function and performance of Woodpecker, it is found that Woodpecker can carry on the information propagation law of social network. Analysis, prediction and control. This article can better understand the law of information transmission of social network, predict the information dissemination situation and explosive behavior of social network effectively, and control the information dissemination of social network carefully. The research results are in information transmission, business marketing, information recommendation, social management and so on. All fields have strong reference value and reference significance.

【学位授予单位】：哈尔滨理工大学
【学位级别】：博士
【学位授予年份】：2016
【分类号】：TP393.09;G206

【相似文献】