Multi-Scale Self-Supervised Graph Contrastive Learning With Injective Node Augmentation
文献类型：期刊论文
| 作者 | Zhang, Haonan1; Ren, Yuyang1; Fu, Luoyi1; Wang, Xinbing1; Chen, Guihai1; Zhou, Chenghu2 |
| 刊名 | IEEE TRANSACTIONS ON KNOWLEDGE AND DATA ENGINEERING
 |
| 出版日期 | 2024 |
| 卷号 | 36期号:1页码:261-274 |
| 关键词 | Graph contrastive learning graph representation learning node augmentation self-supervised learning |
| ISSN号 | 1041-4347 |
| DOI | 10.1109/TKDE.2023.3278463 |
| 通讯作者 | Wang, Xinbing(xwang8@sjtu.edu.cn) |
| 英文摘要 | Graph Contrastive Learning (GCL) with Graph Neural Networks (GNN) has emerged as a promising method for learning latent node representations in a self-supervised manner. Most of existing GCL methods employ random sampling for graph view augmentation and maximize the agreement of the node representations between the views. However, the random augmentation manner, which is likely to produce very similar graph view samplings, may easily result in incomplete nodal contextual information, thus weakening the discrimination of node representations. To this end, this paper proposes a novel trainable scheme from the perspective of node augmentation, which is theoretically proved to be injective and utilizes the subgraphs consisting of each node with its neighbors to enhance the distinguishability of nodal view. Notably, our proposed scheme tries to enrich node representations via a multi-scale contrastive training that integrates three different levels of training granularity, i.e., subgraph level, graph- and node-level contextual information. In particular, the subgraph-level objective between augmented and original node views is constructed to enhance the discrimination of node representations while graph- and node-level objectives with global and local information from the original graph are developed to improve the generalization ability of representations. Experiment results demonstrate that our framework outperforms existing state-of-the-art baselines and even surpasses several supervised counterparts on four real-world datasets for node classification. |
| 资助项目 | NSF China |
| WOS研究方向 | Computer Science ; Engineering |
| 语种 | 英语 |
| WOS记录号 | WOS:001124222100019 |
| 出版者 | IEEE COMPUTER SOC |
| 资助机构 | NSF China |
| 源URL | [http://ir.igsnrr.ac.cn/handle/311030/201704]  |
| 专题 | 中国科学院地理科学与资源研究所 |
| 通讯作者 | Wang, Xinbing |
| 作者单位 | 1.Shanghai Jiao Tong Univ, Sch Elect Informat & Elect Engn, Shanghai 200240, Peoples R China 2.Chinese Acad Sci, Inst Geog Sci & Nat Resources Res, Beijing 100045, Peoples R China |
| 推荐引用方式 GB/T 7714 | Zhang, Haonan,Ren, Yuyang,Fu, Luoyi,et al. Multi-Scale Self-Supervised Graph Contrastive Learning With Injective Node Augmentation[J]. IEEE TRANSACTIONS ON KNOWLEDGE AND DATA ENGINEERING,2024,36(1):261-274. |
| APA | Zhang, Haonan,Ren, Yuyang,Fu, Luoyi,Wang, Xinbing,Chen, Guihai,&Zhou, Chenghu.(2024).Multi-Scale Self-Supervised Graph Contrastive Learning With Injective Node Augmentation.IEEE TRANSACTIONS ON KNOWLEDGE AND DATA ENGINEERING,36(1),261-274. |
| MLA | Zhang, Haonan,et al."Multi-Scale Self-Supervised Graph Contrastive Learning With Injective Node Augmentation".IEEE TRANSACTIONS ON KNOWLEDGE AND DATA ENGINEERING 36.1(2024):261-274. |
入库方式： OAI收割
来源：地理科学与资源研究所
浏览0
下载0
收藏0
其他版本
除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。