[1]陈广福,江 玲,韩辉珍.基于节点度异质性惩罚的链路预测方法[J].计算机技术与发展,2022,32(12):81-87.[doi:10. 3969 / j. issn. 1673-629X. 2022. 12. 013]
　CHEN Guang-fu,JIANG Ling,HAN Hui-zhen.Link Prediction Method Based on Node Degree Heterogeneity Penalization[J].,2022,32(12):81-87.[doi:10. 3969 / j. issn. 1673-629X. 2022. 12. 013]
点击复制

基于节点度异质性惩罚的链路预测方法()
分享到：

《计算机技术与发展》[ISSN:1006-6977/CN:61-1281/TN]

卷:
32
期数:
2022年12期
页码:
81-87
栏目:
软件技术与工程
出版日期:
2022-12-10

文章信息/Info

Title:
Link Prediction Method Based on Node Degree Heterogeneity Penalization
文章编号:
1673-629X(2022)12-0081-07
作者:
陈广福1; 2 ; 江 玲1 ; 韩辉珍1
1. 武夷学院 数学与计算机学院,福建 武夷山 353400;
2. 认知计算与智能信息处理福建省高校重点实验室,福建 武夷山 353400
Author(s):
CHEN Guang-fu1; 2 ; JIANG Ling1 ; HAN Hui-zhen1
1. School of Mathematics and Computer,Wuyi University,Wuyishan 353400,China;
2. Key Laboratory of Cognitive Computing and Intelligent Information Processing in Fujian Education Institutions,Wuyishan 353400,China
关键词:
复杂网络; 链路预测; 度异质性; 平均节点聚类系数; 平均最短路径
Keywords:
complex network; link prediction; degree heterogeneity; average node clustering coefficient; average shortest path
分类号:
TP391
DOI:
10. 3969 / j. issn. 1673-629X. 2022. 12. 013
摘要:
针对大部分现存的链路预测方法仅关注规则网络以及偏好连接现象而导致在稀疏网络获得低质量性能,提出一种节点度异质性惩罚的链路预测框架( NDHP) ,该框架最优预测准确度与网络拓扑特征有密切关联。 首先,计算整个网络节点度获得所有节点对的度异质性相似度;其次,采用惩罚节点度较大机制去惩罚度异质性权重较大的节点抑制节点间差异;最后,通过可调参数将平均节点聚类系数和平均最短路径分别和基于度异质性惩罚框架相关联,获取网络结构信息来弥补网络稀疏信息不足,并提出基于节点度异质性惩罚的平均聚类系数指标( NDHP_AC) 和基于节点度异质性惩罚的平均距离指标( NDHP_AD) 。 此外,在 8 个真实无向无权网络上与最近代表性的方法相比较,所提两个指标在预测缺失链接和鲁棒性两方面性能优于基准指标。 尤其在高度稀疏网络中,所提指标的 AUC 和 AUPR 分别最大提高了 15. 3% 和8. 6% 。
Abstract:
Most of existing link prediction methods only focus on regular networks and preferential attachment phenomenon,which resultsin low quality performance in highly sparse networks. We propose a link prediction framework based on node degree heterogeneitypenalty. The optimal prediction accuracy of the framework is closely related to network topology characteristics. Firstly,the node degreeof the whole network is calculated to obtain the degree heterogeneity similarity of the predicted node pairs. Secondly,the mechanism ofpunishing nodes with higher degree of heterogeneity is used to suppress the differences between nodes. Finally, the average nodeclustering coefficient and the average shortest path are associated with the degree heterogeneity based punishment framework respectivelyby the adjustable parameters to obtain the network structure information to make up for the lack of sparse network information,and theNode Degree Heterogeneity Penalization via Average Clustering coefficient ( NDHP _ AC ) and the Node Degree HeterogeneityPenalization via Average Distance ( NDHP_AD) are proposed. In addition,　compared with the most recent representative method,theperformance of the proposed two indicators is better than the benchmark indicators in predicting missing links and robustness on eight realundirected and unweighted networks. Especially in highly sparse networks,the proposed method improves the maximum AUC and AUPRby 15. 3% and 8. 6% ,respectively.

相似文献/References:

[1]李方洁 刘希玉.复杂网络维的测量[J].计算机技术与发展,2010,(04):61.
　LI Fang-jie,LIU Xi-yu.Measuring Dimensions for Complex Networks[J].,2010,(12):61.
[2]李晶晶 王红.用复杂网络理论分析电网及大停电事故[J].计算机技术与发展,2008,(10):247.
　LI Jing-jing,WANG Hong.Analysis on Power Grids and Blackouts with Complex Network Theory[J].,2008,(12):247.
[3]惠伟 王红.复杂网络在城市公交网络中的实证分析[J].计算机技术与发展,2008,(11):217.
　HUI Wei,WANG Hong.Empirical Analysis of Complex Networks in Public Traffic Networks[J].,2008,(12):217.
[4]赵鹏 蔡庆生 王清毅.一种用于文章推荐系统中的用户模型表示方法[J].计算机技术与发展,2007,(01):4.
　ZHAO Peng,CAI Qing-sheng,WANG Qing-yi.A Novel Representation of User Profile in Document Recommendation System[J].,2007,(12):4.
[5]赵鹏 耿焕同 蔡庆生 王清毅.一种基于加权复杂网络特征的K—means聚类算法[J].计算机技术与发展,2007,(09):35.
　ZHAO Peng,GENG Huan-tong,CAI Qing-sheng,et al.A Novel K- means Clustering Algorithm Based on Weighted Complex Networks Feature[J].,2007,(12):35.
[6]顾亦然 谢鸿飞 李金发.移动通信网络中人类行为动力学的研究[J].计算机技术与发展,2010,(09):57.
　GU Yi-ran,XIE Hong-fei,LI Jin-fa.Studies Based on Complex Network and Dynamics of Human Behavior in MC Network[J].,2010,(12):57.
[7]顾亦然 李金发 谢鸿飞.阵发特性影响因素的研究[J].计算机技术与发展,2010,(09):168.
　GU Yi-ran,LI Jin-fa,XIE Hong-fei.Study on Influence Factors of Characteristic of Burst[J].,2010,(12):168.
[8]何明东 熊建斌 李振坤.基于复杂网络的软件开发方法研究[J].计算机技术与发展,2011,(06):59.
　HE Ming-dong,XIONG Jian-bin,LI Zhen-kun.Complex Network-Based Software Development Method Research[J].,2011,(12):59.
[9]王泽洪 闵妍妮 刘名扬 谭韵天.Pub/Sub系统中基于免疫的新型路由算法[J].计算机技术与发展,2012,(02):6.
　WANG Ze-hong,MIN Yan-ni,LIU Ming-yang,et al.A New Immunity-Based Routing Strategy in Pub/Sub System[J].,2012,(12):6.
[10]马兴福 王红 李园园.基于复杂网络的中小企业板股市网络特性分析[J].计算机技术与发展,2012,(04):172.
　MA Xing-fu,WANG Hong,LI Yuan-yuan.Characteristic Analysis of Small and Medium Enterprises Board Stock Market Network Based on Complex Network[J].,2012,(12):172.
[11]李旗旗,徐 敏.链路预测方法与网络结构的相关性[J].计算机技术与发展,2017,27(12):57.[doi:10.3969/ j. issn.1673-629X.2017.12.013]
　LI Qi-qi,XU Min.Correlation between Link Prediction Method and Network Structure[J].,2017,27(12):57.[doi:10.3969/ j. issn.1673-629X.2017.12.013]
[12]陆圣宇,史 军,刘 宝,等.基于转移自洽和偏好连接的链路预测算法研究[J].计算机技术与发展,2019,29(08):18.[doi:10. 3969 / j. issn. 1673-629X. 2019. 08. 004]
　LU Sheng-yu,SHI Jun,LIU Bao,et al.Research on a Link Prediction Algorithm Based on Transferring Similarity and Preferential Attachment[J].,2019,29(12):18.[doi:10. 3969 / j. issn. 1673-629X. 2019. 08. 004]

更新日期/Last Update: 2022-12-10