基于基尼系数的网络结构洞测量
Gini-coefficient-based measurement of structural holes
邓世果 1吴干华 2杨会杰1
作者信息
- 1. 上海理工大学管理学院,上海200093
- 2. 上海理工大学管理学院,上海200093;华南师范大学南海学院,佛山528225
- 折叠
摘要
用结构洞分析网络结构是评价网络结构引起的竞争优势的一种重要方法.本文提出了一种新的结构洞定量描述方法,即基于主成分分析(PCA)的结构洞的测量.给出了类似于基尼系数的网络贡献度的概念,在满足度分布不变的条件下利用贡献度分析对比无标度网络与对应的随机网络,证实用贡献度评价网络有效性的优越性.得出结论:贡献度S越大,说明网络中每个节点的重要性越平均,结构洞程度越大；反之,贡献度S越小,则说明网络结构信息主要集中在部分节点上,结构洞程度越小.
Abstract
Structural hole is an important concept to estimate pattern-induced competitive power for each node in a social network. Principal component analysis(PCA)was proposed to evaluate the structural holes in a global way. Gini-coefficient for social network was defined to measure quantitatively the structural holes. As examples, the Gini coefficients for some real-world networks and the extended Barabasi-Albert scale-free networks were calculated. It is concluded that the bigger the contribution of S in the network, the more uniform the importance of modes and the greater the degree of structural holes;Conversely,the smaller the contribution of S,showing the network structure information mainly concentrated in part of nodes, the smaller the degree of structural hole.
关键词
结构洞/主成分分析/网络贡献度/复杂网络Key words
structural hole/principal component analysis/network contribution/complex network引用本文复制引用
基金项目
国家自然科学基金(10975099)
国家自然科学基金(10635040)
上海市重点学科建设项目(S30501)
出版年
2011
NETL
NSTL
维普
万方数据