Yuwen Zhao , Weifang Liang , Zhi Gong , Shibing Sun , Shayan Nejadshamsi
{"title":"DSEAGC: Dual-spectral embedding for attributed graph clustering","authors":"Yuwen Zhao , Weifang Liang , Zhi Gong , Shibing Sun , Shayan Nejadshamsi","doi":"10.1016/j.iot.2025.101651","DOIUrl":null,"url":null,"abstract":"<div><div>Community detection in node-attributed networks, where nodes are characterized by both structural connections and attribute information, is a crucial task in network analysis. Accurately identifying these communities reveals underlying patterns and relationships, offering deeper insights into the network’s structure and behavior. However, effectively integrating structural and attribute data remains a challenge. To address this, we introduce a novel framework called DSEAGC (Dual-Spectral Embedding for Attributed Graph Clustering), which jointly leverages structure and attribute spaces through a three-stage spectral learning pipeline. Specifically, DSEAGC constructs separate Laplacian matrices for both graph topology and node attributes, performs independent spectral embeddings, and fuses these views using an adaptive objective function. An iterative optimization technique balances the preservation of structural and attribute information while achieving discrete cluster assignments via spectral rotation. Unlike existing methods, DSEAGC incorporates a dual preservation mechanism and rotation-based discretization to enhance cluster separability. This unified representation enhances community detection by reflecting structural and attribute similarities and capturing both complementary and consensus information. These improvements are evidenced through comprehensive evaluations using multiple clustering metrics, validating the effectiveness and scalability of DSEAGC in practical attributed network scenarios.</div></div>","PeriodicalId":29968,"journal":{"name":"Internet of Things","volume":"32 ","pages":"Article 101651"},"PeriodicalIF":6.0000,"publicationDate":"2025-05-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Internet of Things","FirstCategoryId":"94","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S2542660525001659","RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"COMPUTER SCIENCE, INFORMATION SYSTEMS","Score":null,"Total":0}
引用次数: 0
Abstract
Community detection in node-attributed networks, where nodes are characterized by both structural connections and attribute information, is a crucial task in network analysis. Accurately identifying these communities reveals underlying patterns and relationships, offering deeper insights into the network’s structure and behavior. However, effectively integrating structural and attribute data remains a challenge. To address this, we introduce a novel framework called DSEAGC (Dual-Spectral Embedding for Attributed Graph Clustering), which jointly leverages structure and attribute spaces through a three-stage spectral learning pipeline. Specifically, DSEAGC constructs separate Laplacian matrices for both graph topology and node attributes, performs independent spectral embeddings, and fuses these views using an adaptive objective function. An iterative optimization technique balances the preservation of structural and attribute information while achieving discrete cluster assignments via spectral rotation. Unlike existing methods, DSEAGC incorporates a dual preservation mechanism and rotation-based discretization to enhance cluster separability. This unified representation enhances community detection by reflecting structural and attribute similarities and capturing both complementary and consensus information. These improvements are evidenced through comprehensive evaluations using multiple clustering metrics, validating the effectiveness and scalability of DSEAGC in practical attributed network scenarios.
期刊介绍:
Internet of Things; Engineering Cyber Physical Human Systems is a comprehensive journal encouraging cross collaboration between researchers, engineers and practitioners in the field of IoT & Cyber Physical Human Systems. The journal offers a unique platform to exchange scientific information on the entire breadth of technology, science, and societal applications of the IoT.
The journal will place a high priority on timely publication, and provide a home for high quality.
Furthermore, IOT is interested in publishing topical Special Issues on any aspect of IOT.