项目名称: 移动社会网络的动态社区发现及其信任评价机理研究
项目编号: No.61502163
项目类型: 青年科学基金项目
立项/批准年度: 2016
项目学科: 自动化技术、计算机技术
项目作者: 陈淑红
作者单位: 湖南工程学院
项目金额: 20万元
中文摘要: 移动性、异构性、动态性、社区性、大规模性等特性使得信任问题成为移动社会网络急需解决的关键问题。本项目首先研究基于隐式社会行为图构建高效的能同时检测分层、重叠等各种复杂社区结构的MSN(Mobile Social Networks)社区发现算法及动态更新算法;同时提出基于检测的复杂社区结构构建MSN信任关系评价体系,研究基于二型模糊逻辑的多维度信任评估控制方法,利用二型模糊逻辑-隐马尔科夫模型解决第三方非诚实信任推荐问题。最后,将搭建实验平台,挖掘真实的MSN数据集,对设计的模型和方案进行测试和评估。本项目的研究成果预期将克服目前动态社区发现和信任计算技术存在的局限,同时为自适应MSN可信通信与协议的设计及其安全性能改进提供重要的理论分析基础和依据。
中文关键词: 隐式社会行为图;动态社区发现;二型模糊逻辑;多维度信任评估;信任推荐
英文摘要: Trust evaluation has become a key challenge of mobile social networks caused by its special characteristics, such as mobility, heterogeneity, dynamics, community structure and large-scale properties. This project propose an efficient community structure detection algorithm based on implicit social behavioral graph which can detect hierarchical structure, overlapping structure and various complex community structure simultaneously, and we also propose the corresponding dynamic updating algorithm. Then, based on the detected community structure, we will propose a MSN trust evaluation system. We propose a multidimensional trust evaluation control method based on type-II fuzzy logic. To prevent the dishonest trust recommendation of third parties, we propose adopting type-II fuzzy logic and Hidden Markov Model to solve this problem. Finally, we will develop a prototype system to evaluate and justify the proposed methods on real dataset. These modeling techniques will overcome the limitations of existing dynamic community structure detection trust evaluation techniques. These research results are the theoretic basis for improving the security performances of MSNs and self-adaptive trusted communication protocols.
英文关键词: Implicit Social Behavioral Graph;Dynamic Community Structure Detection;Type-II Fuzzy Logic;Multidimensional Trust Evaluation;Trust Recommendation