Page 221 - 《软件学报》2026年第2期

软件学报 ISSN 1000-9825, CODEN RUXUEW E-mail: jos@iscas.ac.cn
2026,37(2):700−715 [doi: 10.13328/j.cnki.jos.007439] [CSTR: 32375.14.jos.007439] http://www.jos.org.cn
©中国科学院软件研究所版权所有. Tel: +86-10-62562563

*
基于结构关系建模的自监督图表示学习

秦者云 1 , 卢宪凯 1 , 聂秀山 2 , 尹义龙 1

1
(山东大学 软件学院, 山东 济南 250101)
2
(山东建筑大学 计算机科学与技术学院, 山东 济南 250101)
通信作者: 卢宪凯, E-mail: carrierlxk@gmail.com; 尹义龙, E-mail: ylyin@sdu.edu.cn

摘 要: 研究目标是从未标记的图数据中学习健壮的图表示. 开发了一种结构关系建模 (structural relation
modeling, SRM) 框架, 用于自监督图表示学习, 缓解了由未标记数据和图拓扑不平衡引起的固有限制. 首先, 与大
多数现有方法专注于局部结构或节点嵌入不同, 通过在统一框架内对节点、子图和整个图之间的复杂关系 (即局
部-全局关系和节点相关性) 进行建模来捕捉图结构. 这有助于更好地理解图的拓扑结构, 并利用结构自监督信号.
其次, 引入了一种基于分区的子图采样机制, 通过小批量训练缓解了由图拓扑不平衡引起的过度聚合和拓扑衰减.
该机制确保更均匀的信息传播. 第三, 施加了一种节点正则化策略, 以提高训练的稳定性和效率, 产生更精确的结
构表示. 对 12 个公共数据集进行的节点和图分类的广泛实验证明了所提方法的有效性和普适性.
关键词: 自监督图表示学习; 图拓扑不平衡; 结构关系建模; 图结构; 子图
中图法分类号: TP18

中文引用格式: 秦者云, 卢宪凯, 聂秀山, 尹义龙. 基于结构关系建模的自监督图表示学习. 软件学报, 2026, 37(2): 700–715. http://
www.jos.org.cn/1000-9825/7439.htm
英文引用格式: Qin ZY, Lu XK, Nie XS, Yin YL. Self-supervised Graph Representation Learning via Structural Relation Modeling.
Ruan Jian Xue Bao/Journal of Software, 2026, 37(2): 700–715 (in Chinese). http://www.jos.org.cn/1000-9825/7439.htm

Self-supervised Graph Representation Learning via Structural Relation Modeling
1
1
2
QIN Zhe-Yun , LU Xian-Kai , NIE Xiu-Shan , YIN Yi-Long 1
1
(School of Software, Shandong University, Jinan 250101, China)
2
(School of Computer Science and Technology, Shandong Jianzhu University, Jinan 250101, China)
Abstract: This study aims to learn robust graph representations from unlabeled graph data. A novel framework, termed structural relation
modeling (SRM), is proposed for self-supervised graph representation learning to alleviate inherent limitations caused by unlabeled data
and graph topological imbalances. First, rather than focusing solely on local structures or node embeddings as in most existing methods,
this study models complex structural relations, such as local-global relations and node correlations, among nodes, subgraphs, and entire
graphs within a unified framework to better capture graph topology and utilize structural self-supervision. Second, a partition-based
subgraph sampling mechanism is introduced to mitigate over-aggregation and topological decay induced by graph topological imbalance,
enabling more uniform information propagation through mini-batch training. Third, a node regularization strategy is applied to improve
training stability and efficiency, resulting in more accurate structural representations. Extensive experiments on node and graph
classification across 12 public datasets demonstrate the effectiveness and generalizability of the proposed method.
Key words: self-supervised graph representation learning; graph topological imbalance; structural relation modeling (SRM); graph structure;
subgraph

* 基金项目: 国家自然科学基金 (U23A20389, 62176139, 62106128, 62176141)
收稿时间: 2024-03-13; 修改时间: 2024-08-02; 采用时间: 2025-03-25; jos 在线出版时间: 2025-11-05
CNKI 网络首发时间: 2025-11-06