基于K-shell分解与邻居节点度去噪的链路预测方法  

Link prediction method based on K-shell decomposition and neighbor node degree denoising

作　　者：张希康 李泽滔[1] Zhang Xikang;Li Zetao(School of Electrical Engineering,Guizhou University,Guiyang 550025,China)

机构地区：[1]贵州大学电气工程学院,贵阳550025

出　　处：《计算机应用研究》2022年第11期3270-3274,共5页Application Research of Computers

基　　金：国家自然科学基金资助项目(61963009);贵州省科技计划项目(黔科合支撑[2019]2154号)。

摘　　要：链路预测是研究复杂网络结构和演化机制的重要工具,提高链路预测的精度具有重要价值。针对传统的基于网络拓扑结构相似性算法预测精度偏低的问题,从网络优化去噪的角度进行分析,提出了一种基于K-shell分解与邻居节点度(KSDNN)去噪的链路预测方法。该方法首先从全局的角度通过K-shell分解对复杂网络中所有节点进行重要性排序,然后从局部的角度结合节点邻居节点的度对节点重要性进行综合评判,最后对网络数据进行优化后进行链路预测。通过在四个不同的真实网络进行验证,实验结果表明,所提方法预测精度优于K-shell去噪的方法,且相较于传统算法预测精度平均提升了2%左右。Link prediction is an important tool to study the structure and evolution mechanism of complex networks,and it is of great value to improve the accuracy of link prediction.To address the problem of low prediction accuracy of traditional network topology similarity-based algorithms,this paper proposed a link prediction method based on K-shell decomposition with neighbor node degree(KSDNN) denoising from the perspective of network optimization denoising.The method firstly ranked the importance of all nodes in a complex network by K-shell decomposition from a global perspective,then made a comprehensive evaluation of node importance from a local perspective by combining the degrees of nodes’ neighbor nodes,and finally performed link prediction after optimizing the network data.Through verification on four different real networks,the experimental results show that the prediction accuracy of the proposed method is better than that of the K-shell denoising method,and the prediction accuracy is improved by about 2% on average compared with the traditional algorithm.

关 键 词：链路预测 复杂网络 K-shell分解 邻居节点度 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]