摘要
为了逼近真实世界复杂网络,方向性和模糊性是不断发展的网络社团提取方法必须加以考虑的重要特性。在有向网络中研究和分析模糊社团结构具有一定的现实意义。为开发有向网络的模糊社团分析方法,本文构建了一种基于改进的非负奇异值分解法的最优化框架,用矩阵分解技术得到了有向网络中描述点团之间有向关联关系的一种新模糊度量。实验证明新度量不仅能用来提取有向网络的模糊社团,而且可用于对有向网络的社团拓扑结构做深度分析,例如分析社团之间的宏观有向关联关系,对社团间的有向连接贡献程度加以量化并提取重要节点等。城市路网含有大量方向性信息,路网效能强烈依赖其结构鲁棒性,本方法可用于城市路网社团结构划分与脆弱性分析,为提高路网效能提供整体结构特性方面的参考。
It is of practical value to study and analyze the structure of fuzzy communities in directed network.In order to approach the real world complex network,directivity and fuzziness are important characteristics that must be considered in the methods of the network community extraction.To develop fuzzy analysis method in directed network,we introduce an optimal framework of Non-negative Singular Value Decomposition.By matrix decomposition approach,we obtained a new fuzzy metric characterizing the directed correlation degree between node and community in a directional manner.Experiments prove that the new metric can be used not only to extract fuzzy communities in directed networks,but also to analyze the topology of the directed network,such as analysing the directed association between communities or extracting the important nodes.This method can also provide means for the study of topological properties of road network.
作者
石发进
陈金章
赵昆
黄鹤
李炜光
SHI Fa-jin;CHEN Jin-zhang;ZHAO Kun;HUANG He;LI Wei-guang(Zhongjiao Tongli Construction CO.,Ltd.,Xi’an 710075,China;Key Laboratory for Special Area Highway Engineering of Ministry of Education,Chang’an University2,Xi’an 710064,China)
出处
《电子设计工程》
2019年第3期20-25,共6页
Electronic Design Engineering
基金
国家自然科学基金青年科学基金项目(41101357)
关键词
有向网络
模糊聚类
非对称特征矩阵
有向关联度
directed network
fuzzy clustering
asymmetric eigenmatrix
directed correlation degree
