基于无标度性和资源分配的网络链路预测算法  

Link prediction algorithm of network based on the scale-free and the resource allocation

作　　者：刘凯 李晨璞[2] 邹龙 王海龙 王浩森 LIU Kai;LI Chenpu;ZOU Long;WANG Hailong;WANG Haosen(School of Information Engineering,Hebei University of Architecture,Zhangjiakou 075000,China;Department of Mathematics and Physics,Hebei University of Architecture,Zhangjiakou 075000,China)

机构地区：[1]河北建筑工程学院信息工程学院,河北张家口075000 [2]河北建筑工程学院数理系,河北张家口075000

出　　处：《高师理科学刊》2023年第4期19-26,共8页Journal of Science of Teachers＇College and University

基　　金：河北省自然科学基金项目(D2022404003);河北省高等学校科学研究计划拔尖项目(BJ2021054);河北建筑工程学院功能材料与结构力学团队项目(TD202011);河北省高等教育教学改革研究与实践项目(2018GJJG328)。

摘　　要：链路预测是指根据网络结构中的已知信息来推测网络产生新链接的可能性.资源分配算法(RA)是一种简单高效的基于网络局部信息的链路预测算法,但RA算法中并未考虑被预测两节点自身的信息.在RA算法基础上考虑被预测节点间的链接偏好(PA),提出基于网络结构的资源分配RSF和RSW算法.在14个实际网络中进行链路预测实验,验证了所给算法的有效性和鲁棒性,并分析了网络结构的无标度性对算法预测性能的影响.Link prediction refers to inferring the possibility of a network generating new links based on known information in the network structure.Resource allocation algorithm(RA)is a simple and efficient link prediction algorithm based on network local information,however,the information of the two nodes that to be predicted is not be considered in the RA algorithm.Based on the RA algorithm,the link preference(PA)between the predicted nodes is considered,and resource allocation RSF and RSW algorithms are proposed based on network structure.Link prediction experiments were conducted on 14 actual networks to verify the effectiveness and robustness of the algorithm,and the impact of scale-free network structure on the algorithm′s prediction performance was analyzed.

关 键 词：复杂网络 链路预测 节点度 无标度性 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]