TY - GEN
T1 - Disentangled Hierarchical Attention Graph Neural Network for Recommendation
AU - He, Weijie
AU - Ouyang, Yuanxin
AU - Peng, Keqin
AU - Rong, Wenge
AU - Xiong, Zhang
N1 - Publisher Copyright:
© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024.
PY - 2024
Y1 - 2024
N2 - Heterogeneous information networks (HIN) have been widely used in recommendation systems, aiming to solve how to model complex interactions between entities and data sparsity issue. Due to the excellent performance of Graph Neural Networks (GNN) in representation learning, they are applied in recommendation systems based on HIN. However, most current works focusing on HIN overlook the entanglement of latent factors originating from different aspects. Besides, most of them use meta path-based methods, which fail to consider the semantic information among the paths. In this paper, we propose a Disentangled Hierarchical Attention Graph Neural Network for Recommendation (DHARec), which applies disentangled representations for nodes in HIN. Instead of relying solely on meta paths, we leverage one-hop semantic relation neighbors to aggregate representations based on hierarchical attention, including intra relation and inter relation attention. Specifically, intra relation attention is primarily used to learn the contribution of a neighbor within the same semantic relation, while inter relation attention focuses on learning the importance of different semantic relations and fusing representations from these relations with appropriate weights. Extensive experimental results on three HIN-based datasets demonstrate that our approach outperforms existing methods.
AB - Heterogeneous information networks (HIN) have been widely used in recommendation systems, aiming to solve how to model complex interactions between entities and data sparsity issue. Due to the excellent performance of Graph Neural Networks (GNN) in representation learning, they are applied in recommendation systems based on HIN. However, most current works focusing on HIN overlook the entanglement of latent factors originating from different aspects. Besides, most of them use meta path-based methods, which fail to consider the semantic information among the paths. In this paper, we propose a Disentangled Hierarchical Attention Graph Neural Network for Recommendation (DHARec), which applies disentangled representations for nodes in HIN. Instead of relying solely on meta paths, we leverage one-hop semantic relation neighbors to aggregate representations based on hierarchical attention, including intra relation and inter relation attention. Specifically, intra relation attention is primarily used to learn the contribution of a neighbor within the same semantic relation, while inter relation attention focuses on learning the importance of different semantic relations and fusing representations from these relations with appropriate weights. Extensive experimental results on three HIN-based datasets demonstrate that our approach outperforms existing methods.
KW - Disentangled Representation Learning
KW - Heterogeneous Information Networks
KW - Hierarchical Attention
KW - Top-N Recommendation
UR - https://www.scopus.com/pages/publications/85201225908
U2 - 10.1007/978-981-97-5663-6_35
DO - 10.1007/978-981-97-5663-6_35
M3 - 会议稿件
AN - SCOPUS:85201225908
SN - 9789819756629
T3 - Lecture Notes in Computer Science
SP - 415
EP - 426
BT - Advanced Intelligent Computing Technology and Applications - 20th International Conference, ICIC 2024, Proceedings
A2 - Huang, De-Shuang
A2 - Zhang, Qinhu
A2 - Zhang, Xiankun
PB - Springer Science and Business Media Deutschland GmbH
T2 - 20th International Conference on Intelligent Computing, ICIC 2024
Y2 - 5 August 2024 through 8 August 2024
ER -