《计算机应用》唯一官方网站 ›› 2026, Vol. 46 ›› Issue (8): 2421-2431.DOI: 10.11772/j.issn.1001-9081.2026010069

• 人工智能 • 上一篇    下一篇

基于Wasserstein正则化增强的图表示学习

袁立宁1,2()   

  1. 1.中国人民公安大学 国家安全学院,北京 100038
    2.广西警察学院 信息技术学院,南宁 530028
  • 收稿日期:2026-01-26 修回日期:2026-04-22 接受日期:2026-04-22 发布日期:2026-04-30 出版日期:2026-08-10
  • 通讯作者: 袁立宁
  • 作者简介:袁立宁(1995—),男,河北唐山人,高级工程师,博士研究生,CCF会员,主要研究方向:情报学、图表示学习。
  • 基金资助:
    国家重点研发计划项目(2023YFC3321604);广西重点研发计划项目(桂科AB25069263);广西重点研发计划项目(桂科AB22035034);中国人民公安大学研究生科研创新项目(2025YJSKY012)

Graph representation learning enhanced by Wasserstein regularization

Lining YUAN1,2()   

  1. 1.School of National Security,People’s Public Security University of China,Beijing 100038,China
    2.School of Information Technology,Guangxi Police College,Nanning Guangxi 530028,China
  • Received:2026-01-26 Revised:2026-04-22 Accepted:2026-04-22 Online:2026-04-30 Published:2026-08-10
  • Contact: Lining YUAN
  • Supported by:
    National Key Research and Development Program of China(2023YFC3321604);Postgraduate Scientific Research Innovation Project of People’s Public Security University of China(2025YJSKY012);Key Research and Development Program of Guangxi(GuiKeAB25069263)

摘要:

概率化图表示学习(GRL)模型常依赖KL(Kullback-Leibler)散度约束节点表示的潜在分布,存在后验坍塌、支撑集敏感以及训练梯度不稳定等问题。为此,提出一种基于Wasserstein距离的GRL模型——Wasserstein图自编码器(WGAE),将二阶Wasserstein距离的闭式解作为正则化项,从最优传输理论视角约束潜在分布。相较于KL散度,Wasserstein距离的优势是支撑集鲁棒性与梯度连续性,使概率化GRL模型能避免因后验坍塌导致的表示退化,鼓励模型学习几何结构更清晰且维度间更解纠缠的潜在表示,并优化训练动态。实验结果表明,WGAE不仅提升了下游任务的预测性能,还改善了表示质量与训练动态。通过基准图数据集的多项实验全面验证了所提模型在预测精度、训练稳定性、信息编码效率、抗数据扰动以及嵌入空间解纠缠等方面的优势。可见,Wasserstein正则化能够与多种先进的GRL机制互补增强并带来稳定的性能增益,为构建更稳健且更高效的GRL优化组件提供现实依据。

关键词: 图表示学习, 变分图自编码器, KL散度, Wasserstein距离, Wasserstein图自编码器

Abstract:

Probabilistic Graph Representation Learning (GRL) models typically rely on KL (Kullback-Leibler) divergence to constrain latent distribution of node representations, and suffer from posterior collapse, support set sensitivity, and unstable training gradient. Therefore, we proposed a GRL model based on Wasserstein distance — Wasserstein Graph AutoEncoder (WGAE) to use the closed-form solution of second-order Wasserstein distance as a regularization term to constrain latent distribution from the perspective of optimal transport theory. Compared to KL divergence, the advantage of Wasserstein distance lies in the robustness of support set and gradient continuity, which enables probabilistic graph representation learning models to avoid representation degradation caused by posterior collapse, encourages models to learn potential representations with clearer geometric structure and more disentangled dimensions, and optimize training dynamics. Experimental results demonstrate that WGAE not only improves the prediction performance of graph representation learning models in downstream tasks, but also enhances representation quality and training dynamics. At the same time, through multiple experiments on benchmark graph datasets, the proposed model’s advantages in prediction accuracy, training stability, information encoding efficiency, resistance to data perturbations, and embedding space entanglement resolution are verified comprehensively. It can be seen that Wasserstein regularization can complement and enhance various advanced GRL mechanisms, bringing stable performance gains, and providing a practical basis for building more robust and efficient GRL optimization components.

Key words: Graph Representation Learning (GRL), Variational Graph AutoEncoder (VGAE), KL (Kullback-Leibler) divergence, Wasserstein distance, Wasserstein Graph AutoEncoder (WGAE)

中图分类号: