基于网络局部和全局特性的组合中心性度量措施

阐述了复杂网络中节点的中心性(即节点的重要性)对网络鲁棒性的重大影响,评估节点的多种重要性方法各自的优点与局限性.结合逆和指数ISI、度中心性DC以及介数中心性BC提出一种基于两种人工网络和两种真实网络的组合中心性度量方法IDB,利用删除节点前后网络的最大连通子图的变化对节点的重要性进行刻画仿真实验,验证了该方法的可行性和有效性.仿真结果表明,提出的组合中心性度量方法在节点重要性排序性能优于单一节点重要性排序性能.

Measures of combinatorial centrality based on network local and global characteristics

This paper describes the significant influence of node centrality (i.e., the importance of nodes) on network robustness in complex networks, and elaborates also advantages as well as limitations of various importance methods for evaluating nodes. Combining inverse sum index(ISI), degree centrality(DC)and betweenness centrality(BC), a measurement method of combinatorial centrality(IDB)is proposed based on two kinds of artificial networks and two real networks. Simulation experiments are carried out to characterize the importance of nodes by differences of the maximum connected sub-graph of the network before and after the nodes are deleted. The feasibility and effectiveness of this method are verified by the simulation. The simulation results show that the performance of the combined centrality measurement method proposed is better than that of a single node in the aspect of ranking of node importance.