李星,
兰巨龙,
卫红权,
刘树新,
国家数字交换系统工程技术研究中心 郑州 450002
基金项目:国家自然科学基金(61803384),国家自然科学基金创新研究群体项目(61521003)
详细信息
作者简介:王凯:男,1980年生,副研究员,博士生,研究方向为链路预测、社会网络分析
李星:男,1987年生,助理研究员,博士生,研究方向为链路预测
兰巨龙:男,1962年生,教授,博士生导师,研究方向为新型网络体系,网络动力学
卫红权:男,1970年生,副研究员,硕士生导师,研究方向为社团发现
刘树新:男,1987年生,助理研究员,博士,研究方向为复杂网络演化、链路预测
通讯作者:刘树新 liushuxin11@126.com, liushuxin11@gmail.com
中图分类号:TN915, TP391计量
文章访问数:1461
HTML全文浏览量:735
PDF下载量:62
被引次数:0
出版历程
收稿日期:2019-05-13
修回日期:2019-09-10
网络出版日期:2019-09-19
刊出日期:2020-03-19
A New Link Prediction Method for Complex Networks Based onTopological Effectiveness of Resource Transmission Paths
Kai WANG,Xing LI,
Julong LAN,
Hongquan WEI,
Shuxin LIU,
National Digital Switching System Engineering and Technological R&D Center, Zhengzhou 450002, China
Funds:The National Natural Science Foundation of China (61803384), The National Natural Science Foundation Innovation Research Group Project of China (61521003)
摘要
摘要:链路预测旨在利用网络中已有的拓扑结构或其他信息,预测未连边节点间存在连接的可能性。资源分配指标具有较低复杂度的同时取得了较好的预测效果,但在资源传输过程的描述中缺少对路径有效性的刻画。资源传输过程是网络演化连边产生的重要内在动力,通过分析节点间资源传输路径周围拓扑的有效性,该文提出一种基于资源传输路径有效性的链路预测方法。该方法首先分析了节点间潜在的资源传输路径对资源传输量的影响,提出资源传输路径有效性的量化方法。然后,基于资源传输路径的有效性,通过对双向资源传输量进行刻画,提出了节点间传输路径的有效性指标。在12个实际网络数据集上的实验测试表明,相比其他基于相似性的链路预测方法,该方法在AUC和Precision衡量标准下能够取得更好的效果。
关键词:复杂网络/
链路预测/
资源传输路径/
有效性
Abstract:Link prediction considers to discover the unknown or missing links of complex networks by using the existing topology or other information. Resource Allocation index can achieve a good performance with low complexity. However, it ignores the path effectiveness of resource transmission process. The resource transmission process is an important internal driving force for the evolution of the network. By analyzing the effectiveness of the topology around the resource transmission path between nodes, a link prediction method based on topological effectiveness of resource transmission paths is proposed. Firstly, the influence of potential resource transmission paths between nodes on resource transmission is analyzed, and a quantitative method for resource transmission path effectiveness is proposed. Then, based on the effectiveness of the resource transmission path, after studying the two-way resource transmission amount between two nodes, the transmission path effectiveness index is proposed. The experimental results of 12 real networks show that compared with other link prediction methods, the proposed method can achieve higher prediction accuracy under the AUC and Precision metrics.
Key words:Complex network/
Link prediction/
Resource transmission path/
Effectiveness
PDF全文下载地址:
https://jeit.ac.cn/article/exportPdf?id=d4f10ed8-1d91-4865-9c17-02e5c22f4f9b