项目名称: 面向移动社交网络的上下文数据辅助的社区结构研究
项目编号: No.61472024
项目类型: 面上项目
立项/批准年度: 2015
项目学科: 计算机科学学科
项目作者: 童超
作者单位: 北京航空航天大学
项目金额: 81万元
中文摘要: 移动社交网络的社区结构研究具有重要的理论意义、研究价值和广阔的应用前景。本项目拟结合移动社交网络的上下文数据对社区结构发现和演化进行研究。具体包括:1)设计移动社交网络上下文数据的采集分析系统,对采集到的上下文数据建立语义表示和关联模型,生成具备时空特征的移动社交网络;2)提出保持社区结构的采样算法,对网络进行科学合理地采样,采样结果能够很好地维持社区结构;3)针对后续提出的社区发现算法需要知道社区数量等先验知识的问题,通过理论分析与证明提出基于Laplacian矩阵Jordan型的社区数量与社区骨干结构等先验知识发现方法;4)在上下文数据的语义表示、关联模型和先验知识的基础上,提出基于上下文数据的移动社交网络重叠社区发现算法,并利用拟提出的符合重叠社区特征的社区发现算法评价指标进行评价;5)结合宏观和微观角度进行分析、归纳社区演化规律,提出符合移动社交网络生长规律的社区演化模型。
中文关键词: 移动社交网络;社区结构;社区发现;移动计算;社会网络分析
英文摘要: The research on mobile social networks community structure has important theory significance, research value and broad application prospects. We will research the community detection and community evolution baed on the context data of mobile social networks. The research work includes: 1) We will design the mobile social networks context data collecting and analyzing system, establish the semantic representation and correlation model for these data, and get the abstract mobile social networks with spatiotemporal attribute; 2) We will propose some sampling algorithms which can keep the network community structure well, sample the network data scientifically and reasonably; 3) Some community detection algorithms need to know priori knowledge, for instance, the number of communities or the communities backbone structure, we plan to propose methods for finding such prior knowledge beaded on Laplacian matrices Jordan forms of the network through theoretical analysis and proof; 4) Based on the priori knowledge, semantic representations and correlation models for context data, we will propose some overlapping community detection algorithms based on context data, and give a serious of evaluation indexes to evaluate these algorithms; 5) For the evolution of community, we intend to study the evolution process in both microscopic and macroscopic manners, and present some models that can depict the evolution of community in real mobile social networks.
英文关键词: Mobile Social Networks;Community Struture;Communitity Detection;Mobile Computing;Social Networks Analysis