社交网络中影响最大化的研究与应用

发布时间：2018-01-05 05:35

本文关键词：社交网络中影响最大化的研究与应用　出处：《重庆大学》2015年硕士论文　论文类型：学位论文

【摘要】：在现代互联网环境下,人们逐渐习惯于社交网络生活,并将社交网络作为当前信息获取和传播的主要工具。但是,社交网络中的用户联系非常复杂,难以快速追踪信息传播的路径,加上网络实名认证机制的滞后,使得社交网络中经常充斥虚假等负面信息。因此,研究如何利用社交网络的特性进行合理的舆论引导,快速、高效的抑制网络谣言的传播有着重要的应用价值和现实意义。社交网络中个体通过交流、分享和推荐信息等互动活动,从而形成具有某一特点的团体,如何利用这些社区等特性挖掘网络中影响力最大的节点,并通过其间的交互作用,使得网络中的信息传播和扩散影响范围最大,从而形成正面的引导是社交网络研究的一个热点问题。本文采用复杂网络的相关研究技术建立网络中信息传播的模型,利用社区网络的拓扑结构对网络中的节点进行影响力最大化的分析,通过有目标的引导方案实现目标舆情的传播路径的阻断或者变更,从而实现对舆情的发展进行正确的引导。首先,论文对影响最大化的研究进行了概述,引出了影响函数的次模属性,提出基于这个属性一种改进的贪心爬山算法,克服了传播函数在计算整个网络影响力时的时间复杂度过高,以及传播函数的计算仅局限于由若干邻居构成的局部子图时造成的精确性不足两个问题。然后,论文根据社交网络的社区特征能对信息在网络中的传播过程的产生重要影响的特性,建立了网络节点平均影响等若干计算节点影响的参数,提出了基于社区结构的影响最大化的计算模型并通过实验对模型进行了验证分析。最后,论文利用该计算模型设计并实现了基于影响最大化的舆论牵引策略算法,并将之应用到社区舆论引导系统中获得了较好的引导效果。
[Abstract]:In the modern Internet environment, people are gradually used to the social network life, and the social network as the main tool of information acquisition and dissemination, but the social network user contact is very complex. It is difficult to track the path of information dissemination quickly, coupled with the lag of the real name authentication mechanism, which makes social networks often filled with false and other negative information. It has important application value and practical significance to study how to use the characteristics of social network to guide public opinion and restrain the spread of network rumors quickly and efficiently. Sharing and recommending information and other interactive activities, so as to form a group with a certain characteristics, how to use the characteristics of these communities to mine the most influential nodes in the network, and through the interaction between them. Make the network of information dissemination and diffusion of the largest impact. In order to form positive guidance is a hot issue in the research of social networks. This paper uses the related research technology of complex networks to establish the model of information dissemination in the network. Using the topological structure of the community network to maximize the influence of the nodes in the network through the targeted guiding scheme to achieve the blocking or changing of the transmission path of the target public opinion. In order to achieve the correct guidance of the development of public opinion. Firstly, the thesis summarizes the research of the maximum impact, and leads to the submodule attribute of the influence function. An improved greedy mountain climbing algorithm based on this attribute is proposed, which overcomes the time complexity of the propagation function in calculating the influence of the whole network. And the calculation of propagation function is limited to two problems: the accuracy of local subgraph composed of several neighbors is not enough. According to the social network community characteristics can have an important impact on the transmission of information in the network, the paper established a number of computing node impact parameters such as the average impact of network nodes. Proposed the community structure based on the impact maximization calculation model and through the experiment to carry on the verification analysis to the model. Finally. This paper designs and implements a public opinion traction strategy algorithm based on maximization of influence, and applies it to the community public opinion guidance system to obtain a good guiding effect.
【学位授予单位】：重庆大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TP393.09

【相似文献】