基于节点度和连接方式的网络异质性测度方法

Method for Measuring the Heterogeneity of Networks Based on the Degrees and Connection Patterns of Nodes

在复杂网络异质性研究中,以往主要基于单一的网络节点度分布和度相对重要程度构建指标来刻画异质性,存在一定的缺陷。对此,综合考虑网络节点度的重要性和网络节点连接的方式,运用邻接矩阵谱理论的思想,提出一种新的网络异质性测度方法并得到网络异质性离差指标。算例表明该指标运算简便,结果合理。因此新方法可以克服传统方法的不足,更为有效地反映网络的结构特征,从而加深对网络复杂性的认识。

In the research on the complex network heterogeneity,the former methods are to construct the index based on the degree distribution and relative degree importance solely to characterize the heterogeneity.But these methods have some shortcomings.In view of this,considering the importance of node degrees and node connection patterns,this paper proposes a new measure method for network heterogeneity by using the spectral theory of adjacency matrix,and obtain an index called network heterogeneity deviation.The illustrative examples show that the index can be computed simply and the results are reasonable.So the new measure method can overcome the shortcomings of the traditional methods,reflect the network structure features more effectively,deepen the understanding of network complexity.