文献类型：期刊文献

英文题名：Heterogeneity-Aware Federated Graph Neural Networks for Incomplete Multi-View Clustering

作者：Yan, Xueming[1,2];Wang, Ziqi[3];Jin, Yaochu[4]

机构：[1]Guangdong Univ Foreign Studies, Sch Informat Sci & Technol, Guangzhou 510420, Peoples R China;[2]Guangdong Engn Res Ctr Data Secur Governance & Pri, Guangzhou 510006, Peoples R China;[3]East China Univ Sci & Technol, Sch Informat Sci & Engn, Shanghai 200237, Peoples R China;[4]Westlake Univ, Sch Engn, Hangzhou 310024, Peoples R China

年份：2026

卷号：10

期号：1

起止页码：369

外文期刊名：IEEE TRANSACTIONS ON EMERGING TOPICS IN COMPUTATIONAL INTELLIGENCE

收录：;EI(收录号：20252718700596);WOS:【SCI-EXPANDED(收录号:WOS:001518811100001)】；

基金：This work was supported in part by the International Collaboration Fund for Creative Research Teams(ICFCRT) of NSFC under Grant W2441019, in part by the National Natural Science Foundation of China under Grant 62136003, in part by the Major Program of the Natural Science Foundation of Zhejiang Province under Grant D25F020001, and in part by the Guangdong Philosophy and Social Science Foundation under Grant GD25YGG26.

语种：英文

外文关键词：Feature extraction; Federated learning; Distributed databases; Autoencoders; Accuracy; Training; Servers; Computational modeling; Computational intelligence; Semantics; Heterogeneous graph neural networks; federated learning; incomplete multi-view clustering

摘要：Federated multi-view clustering aims to collaboratively learn a global clustering model from decentralized and privacy-sensitive data distributed across multiple clients. However, existing approaches face two major challenges: the absence of supervision signals and the heterogeneity across multi-view features, which hinder the extraction of consistent and complementary clustering information. Moreover, the inherent incompleteness of multi-view data in federated scenarios further complicates the learning process. To tackle these issues, we propose a Heterogeneity-aware Federated Graph Neural Networks (HafGNN) for the incomplete multi-view clustering. HafGNN employs heterogeneous graph neural network-based autoencoders for the different clients to capture both view-specific and structural representations while preserving data locality. A server-side aggregation mechanism aligns heterogeneous features from overlapping instances to construct a global latent representation. Additionally, global pseudo-labels are generated to guide view completion and enhance clustering consistency across clients. Extensive experiments on multiple public multi-view datasets demonstrate that HafGNN consistently outperforms state-of-the-art approaches in clustering performance, especially under conditions of view incompleteness and client heterogeneity.