Signal Contrastive Enhanced Graph Collaborative Filtering for Recommendation
Abstract Graph collaborative filtering methods have shown great performance improvements compared with deep neural network-based models. However, these methods suffer from data sparsity and data noise problems. To address these issues, we propose a new contrastive learning-based graph collaborative...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
SpringerOpen
2023-06-01
|
| Series: | Data Science and Engineering |
| Subjects: | |
| Online Access: | https://doi.org/10.1007/s41019-023-00215-w |
| _version_ | 1827808169906143232 |
|---|---|
| author | Zhi-Yuan Li Man-Sheng Chen Yuefang Gao Chang-Dong Wang |
| author_facet | Zhi-Yuan Li Man-Sheng Chen Yuefang Gao Chang-Dong Wang |
| author_sort | Zhi-Yuan Li |
| collection | DOAJ |
| description | Abstract Graph collaborative filtering methods have shown great performance improvements compared with deep neural network-based models. However, these methods suffer from data sparsity and data noise problems. To address these issues, we propose a new contrastive learning-based graph collaborative filtering method to learn more robust representations. The proposed method is called signal contrastive enhanced graph collaborative filtering (SC-GCF), which conducts contrastive learning on graph signals. It has been proved that graph neural networks correspond to low-pass filters on the graph signals from the graph convolution perspective. Different from the previous contrastive learning-based methods, we first pay attention to the diversity of graph signals to directly optimize the informativeness of the graph signals. We introduce a hypergraph module to strengthen the representation learning ability of graph neural networks. The hypergraph learning module utilizes a learnable hypergraph structure to model the latent global dependency relations that graph neural networks cannot depict. Experiments are conducted on four public datasets, and the results show significant improvements compared with the state-of-the-art methods, which confirms the importance of considering signal-level contrastive learning and hypergraph learning. |
| first_indexed | 2024-03-11T22:09:29Z |
| format | Article |
| id | doaj.art-037991f7430c47e3ab6ec0aee7631412 |
| institution | Directory Open Access Journal |
| issn | 2364-1185 2364-1541 |
| language | English |
| last_indexed | 2024-03-11T22:09:29Z |
| publishDate | 2023-06-01 |
| publisher | SpringerOpen |
| record_format | Article |
| series | Data Science and Engineering |
| spelling | doaj.art-037991f7430c47e3ab6ec0aee76314122023-09-24T11:27:29ZengSpringerOpenData Science and Engineering2364-11852364-15412023-06-018331832810.1007/s41019-023-00215-wSignal Contrastive Enhanced Graph Collaborative Filtering for RecommendationZhi-Yuan Li0Man-Sheng Chen1Yuefang Gao2Chang-Dong Wang3School of Computer Science and Engineering, Sun Yat-sen UniversitySchool of Computer Science and Engineering, Sun Yat-sen UniversityCollege of Mathematics and Informatics, South China Agricultural UniversitySchool of Computer Science and Engineering, Sun Yat-sen UniversityAbstract Graph collaborative filtering methods have shown great performance improvements compared with deep neural network-based models. However, these methods suffer from data sparsity and data noise problems. To address these issues, we propose a new contrastive learning-based graph collaborative filtering method to learn more robust representations. The proposed method is called signal contrastive enhanced graph collaborative filtering (SC-GCF), which conducts contrastive learning on graph signals. It has been proved that graph neural networks correspond to low-pass filters on the graph signals from the graph convolution perspective. Different from the previous contrastive learning-based methods, we first pay attention to the diversity of graph signals to directly optimize the informativeness of the graph signals. We introduce a hypergraph module to strengthen the representation learning ability of graph neural networks. The hypergraph learning module utilizes a learnable hypergraph structure to model the latent global dependency relations that graph neural networks cannot depict. Experiments are conducted on four public datasets, and the results show significant improvements compared with the state-of-the-art methods, which confirms the importance of considering signal-level contrastive learning and hypergraph learning.https://doi.org/10.1007/s41019-023-00215-wRecommendationCollaborative filteringContrastive learningGraph signalsHypergraph learning |
| spellingShingle | Zhi-Yuan Li Man-Sheng Chen Yuefang Gao Chang-Dong Wang Signal Contrastive Enhanced Graph Collaborative Filtering for Recommendation Data Science and Engineering Recommendation Collaborative filtering Contrastive learning Graph signals Hypergraph learning |
| title | Signal Contrastive Enhanced Graph Collaborative Filtering for Recommendation |
| title_full | Signal Contrastive Enhanced Graph Collaborative Filtering for Recommendation |
| title_fullStr | Signal Contrastive Enhanced Graph Collaborative Filtering for Recommendation |
| title_full_unstemmed | Signal Contrastive Enhanced Graph Collaborative Filtering for Recommendation |
| title_short | Signal Contrastive Enhanced Graph Collaborative Filtering for Recommendation |
| title_sort | signal contrastive enhanced graph collaborative filtering for recommendation |
| topic | Recommendation Collaborative filtering Contrastive learning Graph signals Hypergraph learning |
| url | https://doi.org/10.1007/s41019-023-00215-w |
| work_keys_str_mv | AT zhiyuanli signalcontrastiveenhancedgraphcollaborativefilteringforrecommendation AT manshengchen signalcontrastiveenhancedgraphcollaborativefilteringforrecommendation AT yuefanggao signalcontrastiveenhancedgraphcollaborativefilteringforrecommendation AT changdongwang signalcontrastiveenhancedgraphcollaborativefilteringforrecommendation |