摘要
节点重要性度量是复杂网络领域的一个重要研究内容,研究者提出了许多中心性度量方法来恒量网络中节点的重要程度,但传统度中心性的定义并没有考虑网络中连边的方向和加权特性,导致其不能有效衡量有向加权网络中节点的重要程度。考虑有向加权网络的特点,提出一种改进的度中心性计算方法--JP-Degree中心性。改进的度中心性综合考虑相邻节点个数、连边的权重以及连边的方向等三个方面的因素得出节点最终的度中心性,能够较为全面地反映三个因素对节点重要性的影响,同时有效避免了单一方向强度过大而导致的节点中心性值很高的情况。在2个真实的社交网络数据上进行实验,结果表明,提出的JP-Degree中心性在衡量节点的连通能力和免疫能力方面有着明显好于传统度中心性的效果。
Evaluation on nodes’ importance is a major direction in the area of complex network,researchers have proposed many centrality measurement methods to evaluate nodes’ importance. But traditional degree centrality cannot accurately estimate the importance of nodes in weighted and directed network as the definition of traditional degree centrality does not consider the directivity and weight of edges. A modified degree centrality named JP-Degree was proposed according to the characteristics of weighted and directed network. The improved degree centrality considered the number of adjacent nodes,the weight of adjacent edges and the direction of adjacent edges comprehensively,which could reflect the influence of three factors on the importance of nodes and effectively eliminate the situation that too large strength from one direction would lead to over-estimated centrality on nodes. Simulation results on two real-world social networks show that JP-Degree centrality has obviously better effects than traditional degree centrality on measuring the connectivity and immune capacity of nodes.
作者
赵构恒
贾鹏
周安民
ZHAO Gouheng;JIA Peng;ZHOU Anmin(College of Cybersecurity,Sichuan University,Chengdu Sichuan 610065,China)
出处
《计算机应用》
CSCD
北大核心
2020年第S01期141-145,共5页
journal of Computer Applications
基金
国家重点研发计划项目(2017YFB0802900)
CCF-启明星辰“鸿雁”科研资助计划项目(CCF-VenustechRP2017002)。
关键词
复杂网络
中心性
有向加权网络
连通性
免疫性
complex network
centrality
directed-weighted network
connectivity
immunity
