基于加权投影的二分网络的链路预测  

LINK PREDICTION OF BIPARTITE NETWORK BASED ON WEIGHTED PROJECTION

作　　者：张佳慧 张婷 吕来水 张娜 Zhang Jiahui;Zhang Ting;Lü Laishui;Zhang Na(School of Computer Science and Engineering,Nanjing University of Science&Technology,Nanjing 210094,Jiangsu,China)

机构地区：[1]南京理工大学计算机科学与工程学院,江苏南京210094

出　　处：《计算机应用与软件》2021年第3期264-268,297,共6页Computer Applications and Software

基　　金：江苏省研究生科研与实践创新计划项目(SJCX18_0149)。

摘　　要：二分网络上的链路预测是复杂网络中的重要研究内容。充分利用拓扑结构,提出一种基于双向投影的二分网络上的链路预测算法。它能同时投影二分网络上的两类节点,分别生成基于顶部节点的投影图和基于底部节点的投影图。通过基于资源分配的加权方法得到同类节点之间的资源权重比,利用相似度度量两类节点之间未链接边资源分配值进行比较,预测出链路边。在实验分析中通过利用AUC来分析算法的准确性和精度,得出该算法具有一定的优势。Link prediction on bipartite network is an important research content in complex network.By making full use of topological structure,this paper puts forward a link prediction algorithm of bipartite network based on bi-directional projection.It could project two classes of nodes on the bipartite network,and generated the projection based on the top node and the projection based on the bottom node respectively.Through the weighted method based on resource allocation,the resource weight ratio between similar nodes was obtained,and we used the similarity measure to compare the resource allocation values of the unlinked edges between the most two types of nodes to predict the link edge.In the experimental analysis,the accuracy and precision of the algorithm are analyzed by using AUC,and it is concluded that this algorithm has certain advantages.

关 键 词：二分网络 链路预测 投影 资源分配 权值 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]