RACE：异构图神经网络的关系级反事实解释

中文摘要 AI 辅助

图神经网络的反事实解释通过识别能使预测翻转的边删除来工作。然而，在异构图上，现有方法首先将图折叠为无类型边，因此无法回答领域专家真正提出的问题：哪种关系类型驱动了该预测？我们提出RACE（关系感知反事实解释），它为每个实例提供精确的逐实例答案。对于每个被解释的实例，对关系子集进行穷举搜索，返回能翻转预测的经过认证的最小关系删除集——或明确报告不存在这样的删除；每个关系级答案随后在属性关系内细化为带类型的边集，并通过单边恢复在离散模型上验证。在冻结骨干网络下，关系级答案是精确且确定性的，而软掩码基线在不同运行（仅随机排序不同）中的成功率波动6-8个百分点。在ACM、源自Cora的图以及ogbn-mag上，RACE相比最强基线将反事实成功率提升最多+2.7个百分点，同时删除更少的边，并在所有数据集上达到所有同任务基线中的最高成功率；该优势在ogbn-arXiv和DBLP上的四个骨干网络中重现，跨种子的关系集一致性高达0.89。一项具有已知生成机制的合成研究证实，搜索能恢复训练模型实际依赖的关系——并在模型未学到任何关系时报告不可行性而非编造归因——因此解释在解释至关重要的地方保持可信。

英文摘要

Counterfactual explanations of graph neural networks identify edge deletions that flip a prediction. On heterogeneous graphs, however, existing methods first collapse the graph into untyped edges, so they cannot answer the question a domain expert actually asks: which relation type drives this prediction? We present RACE (Relation-Aware Counterfactual Explanations), which gives this question an exact, per-instance answer. For every explained instance, an exhaustive search over relation subsets returns the certified minimum relation-deletion set that flips the prediction -- or an explicit report that no such deletion exists; each relation-level answer is then refined into a typed edge set within the attributed relations, verified on the discrete model by single-edge restoration. The relation-level answer is exact and deterministic given the frozen backbone, whereas soft-mask baselines vary by 6-8 pp in success rate across runs differing only in random ordering. On ACM, a Cora-derived graph, and ogbn-mag, RACE improves counterfactual success rate over the strongest baseline by up to +2.7 pp while deleting fewer edges, and attains the highest success rate among all same-task baselines on every dataset; the advantage reproduces across four backbones on ogbn-arXiv and on DBLP, with cross-seed relation-set agreement up to 0.89. A synthetic study with known generating mechanisms confirms that the search recovers the relation the trained model actually relies on -- and reports infeasibility rather than fabricating an attribution when the model has learned none -- so the explanations stay trustworthy exactly where explanations matter.

发表机构

• China Life Insurance Company Ltd.（中国人寿保险股份有限公司）
• Beijing Institute of Technology（北京理工大学）

机构由 AI 辅助整理，请以论文原文为准。