Heterogeneous Graph Neural Network with Adaptive Relation Reconstruction

被引:0
作者
Lin, Weihong [1 ,2 ]
Chen, Zhaoliang [3 ]
Chen, Yuhong [1 ,2 ]
Wang, Shiping [1 ,2 ]
机构
[1] Fuzhou Univ, Coll Comp & Data Sci, Fuzhou 350108, Peoples R China
[2] Fuzhou Univ, Fujian Prov Key Lab Network Comp & Intelligent Inf, Fuzhou 350108, Peoples R China
[3] Hong Kong Baptist Univ, Dept Comp Sci, Hong Kong, Peoples R China
基金
中国国家自然科学基金;
关键词
Graph learning; Heterogeneous information networks; Graph neural networks; Graph augmentation; Semi-supervised classification;
D O I
10.1016/j.neunet.2025.107313
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Topological structures of real-world graphs often exhibit heterogeneity involving diverse nodes and relation types. In recent years, heterogeneous graph learning methods utilizing meta-paths to capture composite relations and guide neighbor selection have garnered considerable attention. However, meta-path based approaches may establish connections between nodes of different categories while overlooking relations between nodes of the same category, decreasing the quality of node embeddings. In light of this, this paper proposes a Heterogeneous Graph Neural Network with Adaptive Relation Reconstruction (HGNN-AR2) that adaptively adjusts the relations to alleviate connection deficiencies and heteromorphic issues. HGNNAR2 is grounded on distinct connections derived from multiple meta-paths. By examining the homomorphic correlations of latent features from each meta-path, we reshape the cross-node connections to explore the pertinent latent relations. Through the relation reconstruction, we unveil unique connections reflected by each meta-path and incorporate them into graph convolutional networks for more comprehensive representations. The proposed model is evaluated on various benchmark heterogeneous graph datasets, demonstrating superior performance compared to state-of-the-art competitors.
引用
收藏
页数:11
相关论文
共 41 条
[1]   Hierarchical Aggregations for High-Dimensional Multiplex Graph Embedding [J].
Abdous, Kamel ;
Mrabah, Nairouz ;
Bouguessa, Mohamed .
IEEE TRANSACTIONS ON KNOWLEDGE AND DATA ENGINEERING, 2024, 36 (04) :1624-1637
[2]  
Bai Q., 2023, P 32 INT C WORLD WID, P523
[3]  
Bengio Y., 2013, arXiv
[4]  
Bo DY, 2022, AAAI CONF ARTIF INTE, P3913
[5]   Heterogeneous Graph Contrastive Learning Network for Personalized Micro-Video Recommendation [J].
Cai, Desheng ;
Qian, Shengsheng ;
Fang, Quan ;
Hu, Jun ;
Ding, Wenkui ;
Xu, Changsheng .
IEEE TRANSACTIONS ON MULTIMEDIA, 2023, 25 :2761-2773
[6]   M 2 ixKG: Mixing for harder negative samples in knowledge graph [J].
Che, Feihu ;
Tao, Jianhua .
NEURAL NETWORKS, 2024, 177
[7]  
Chen Y., 2024, Neural Networks, V183
[8]   Joint learning of feature and topology for multi-view graph convolutional network [J].
Chen, Yuhong ;
Wu, Zhihao ;
Chen, Zhaoliang ;
Dong, Mianxiong ;
Wang, Shiping .
NEURAL NETWORKS, 2023, 168 :161-170
[9]   Attributed Multi-Order Graph Convolutional Network for Heterogeneous Graphs [J].
Chen, Zhaoliang ;
Wu, Zhihao ;
Zhong, Luying ;
Plant, Claudia ;
Wang, Shiping ;
Guo, Wenzhong .
NEURAL NETWORKS, 2024, 174
[10]   Multi-View Graph Convolutional Networks with Differentiable Node Selection [J].
Chen, Zhaoliang ;
Fu, Lele ;
Xiao, Shunxin ;
Wang, Shiping ;
Plant, Claudia ;
Guo, Wenzhong .
ACM TRANSACTIONS ON KNOWLEDGE DISCOVERY FROM DATA, 2024, 18 (01)