Predicting hyperlinks via weighted hypernetwork loop structure

作者全名："Peng, Hao; Li, Shuzhe; Zhao, Dandan; Zhong, Ming; Qian, Cheng; Wang, Wei"

作者地址："[Peng, Hao] Zhejiang Normal Univ, Key Lab Intelligent Educ Technol & Applicat Zhejia, Yingbin Ave, Jinhua 321004, Zhejiang, Peoples R China; [Peng, Hao; Li, Shuzhe; Zhao, Dandan; Zhong, Ming; Qian, Cheng] Zhejiang Normal Univ, Sch Comp Sci & Technol, Yingbin Ave, Jinhua 321004, Zhejiang, Peoples R China; [Wang, Wei] Chongqing Med Univ, Sch Publ Hlth, Chongqing 400016, Peoples R China"

通信作者："Wang, W (通讯作者)，Chongqing Med Univ, Sch Publ Hlth, Chongqing 400016, Peoples R China."

来源：EUROPEAN PHYSICAL JOURNAL-SPECIAL TOPICS

ESI学科分类：PHYSICS

WOS号：WOS:001183600900002

JCR分区：Q2

影响因子：2.6

年份：2024

卷号：　

期号：　

开始页：　

结束页：　

文献类型：Article; Early Access

关键词：　

摘要："Higher-order relationships are prevalent in the real world, and the strength differences between these relationships are often significant and cannot be ignored. Weighted hypernetworks can more accurately represent these prevalent relationships than simple graphs. Due to various limitations, we may not be able to observe all Higher-order relationships that exist within the network. Therefore, algorithms that can perform weighted hyperlink prediction are of great significance. However, current research on link prediction often focuses on simple networks or ordinary higher-order relationships, without considering the important factor of weight. This paper proposes an algorithm based on topologic similarity to predict whether weighted hyperedges are included in weighted hypernetworks, complementing research in this area. On the artificial weighted hypernetwork, the co-author network of various subjects, and the patent Chinese medicine prescription network, we tested the algorithm's robustness and achieved higher accuracy than other methods. Applying this algorithm can handle the prediction of weighted many-to-many relationships and improve accuracy by utilizing the weights on the weighted hypernetwork. Compared to the unweighted version, the algorithm's accuracy on the patent Chinese medicine prescription network has been improved from 97 to 99%. It has been improved from 94 to 96% on the geology co-author network."

基金机构："the National Natural Science Foundation of China [62072412, 61902359, 61702148, 61672468]; National Natural Science Foundation of China [AGK-2018001]; Opening Project of Shanghai Key Laboratory of Integrated Administration Technologies for Information Security [C20607]; Key Lab of Information Network Security, China, Ministry of Public Security, China [cstc2021jcyj-msxmX0132]; Natural Science Foundation of Chongqing, China [W0150]; Program for Youth Innovation in Future Medicine, Chongqing Medical University, China"

基金资助正文："This study was partly supported by the National Natural Science Foundation of China under Grant Nos. 62072412, 61902359, 61702148, and 61672468, the Opening Project of Shanghai Key Laboratory of Integrated Administration Technologies for Information Security under Grant No. AGK-2018001, the Key Lab of Information Network Security, China, Ministry of Public Security, China, under grant no. C20607 and the Natural Science Foundation of Chongqing, China, No. cstc2021jcyj-msxmX0132. Program for Youth Innovation in Future Medicine, Chongqing Medical University, China, No. W0150."