忽略图神经网络的方向性会误导解释结果
Ignoring Directionality Leads to Compromised Graph Neural Network Explanations
- 保留图的有向结构提升解释准确性
- 对称化处理导致解释失真，降低可信度
- 适用于安全关键场景的可解释性研究
图神经网络（GNN）在关键领域中的应用日益广泛，可靠的解释对于支持人类决策至关重要。然而，常见的图对称化处理会丢弃方向信息，造成显著的信息损失并导致误导性解释。我们的分析表明，这种做法会损害解释的保真度。通过理论与实证研究，我们证明保留方向语义能显著提升解释质量，为人类决策者提供更忠实的洞察。这些发现强调了在安全关键应用中采用方向感知的GNN可解释性的重要性。
原文摘要 · Abstract (English)
Graph Neural Networks (GNNs) are increasingly used in critical domains, where reliable explanations are vital for supporting human decision-making. However, the common practice of graph symmetrization discards directional information, leading to significant information loss and misleading explanations. Our analysis demonstrates how this practice compromises explanation fidelity. Through theoretical and empirical studies, we show that preserving directional semantics significantly improves explanation quality, ensuring more faithful insights for human decision-makers. These findings highlight the need for direction-aware GNN explainability in security-critical applications.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。