摘要
引力模型能有效融合节点的多种信息,弥补了传统的节点重要性评估方法考虑的节点信息不全面的问题.然而现有的引力模型相关方法在定义节点质量时考虑的因素单一,并且忽略了邻间拓扑结构在度量节点重要性中的重要作用.针对上述问题,提出一种基于邻居层级分布的引力模型方法来评估节点的重要性,该方法首先将节点的邻居信息和位置信息融合,用来表示引力模型中物体的质量,然后根据节点与其邻域节点的拓扑结构的相似度来定义引力系数,最后利用节点与邻域节点之间的相互作用力之和来度量节点的重要性.在6个真实网络数据集上进行仿真实验,结果表明,该方法的单调性和准确性都优于其他引力模型相关方法.
The gravity model can effectively fuse multiple information of nodes,which make up for the problem of incomplete node information considered by traditional node importance evaluation methods.However,the existing gravity model related methods consider a single factor when defining node mass,and ignore the important role of neighbor topology in measuring node importance.To solve the above problems,a gravity model based on neighborhood hierarchy distribution is proposed for node importance evaluation.Firstly,the neighborhood of nodes and position information are fused to represent the mass of objects in the gravity model.Secondly,the gravity coefficient is defined according to the similarity of the topological structure of the node and neighborhood.Finally,the importance of nodes is measured by the interaction between nodes and neighbor nodes within a given scope.The simulation on six real network datasets shows that the proposed method performs better than other gravity model-related ones in both monotonicity and accuracy.
作者
熊才权
古小惠
吴歆韵
Xiong Caiquan;Gu Xiaohui;Wu Xinyun(School of Computer Science,Hubei University of Technology,Wuhan 430068)
出处
《数学物理学报（A辑）》
CSCD
北大核心
2023年第6期1869-1879,共11页
Acta Mathematica Scientia
基金
国家自然科学基金(61902116)
湖北省科技计划资助项目(2021BLB171)。
关键词
复杂网络
重要节点
引力模型
邻间相互作用
拓扑结构相似性
Complex network
Influential nodes
Gravity model
Neighborhood interaction
Topo-logical similarity.
