分解图数据中的特征、结构和联合信息，提升异常节点检测能力
What Information Matters? Graph Out-of-Distribution Detection via Tri-Component Information Decomposition
- 将图信息拆分为特征、结构和联合三部分，保留与标签相关部分
- 在7个数据集上实现最高34%的FPR95提升，同时保持良好正常数据准确率
- 适合关注图神经网络鲁棒性与异常检测的研究者
图神经网络广泛用于节点分类，但在节点特征或图结构发生分布外（OOD）变化时仍易失效。已有研究表明，标准监督学习方法会捕捉特征或结构中的虚假信号，导致模型脆弱。为此，我们提出TIDE框架，通过三组件信息分解，将信息明确拆分为特征特有、结构特有和联合成分。TIDE旨在保留联合信息中与标签相关的部分，过滤掉虚假的特征与结构信号，从而增强分布内（ID）与分布外（OOD）节点间的区分度。此外，理论与实证分析表明，信息瓶颈目标优于标准监督学习，能提升ID置信度并扩大ID与OOD数据的熵差。在七个数据集上的大量实验验证了TIDE的有效性，相比强基线最高提升34%的FPR95，同时保持竞争性ID准确率。
原文摘要 · Abstract (English)
Graph neural networks are widely used for node classification, but they remain vulnerable to out-of-distribution (OOD) shifts in node features and graph structure. Prior work established that methods trained with standard supervised learning (SL) objectives tend to capture spurious signals from either features and/or structure, leaving the model fragile under distributional changes. To address this, we propose TIDE, a novel and effective Tri-Component Information Decomposition framework that explicitly decomposes information into feature-specific, structure-specific and joint components. TIDE aims to preserve only the label-relevant part of the joint information while filtering out spurious feature- and structure-specific information, thereby enhancing the separation between in-distribution (ID) and OOD nodes. Beyond the framework, we provide theoretical and empirical analyses showing that an information bottleneck objective is preferable to standard SL for graph OOD detection, with higher ID confidence and a greater entropy gap between ID and OOD data. Extensive experiments across seven datasets confirm the efficacy of TIDE, achieving up to a 34% improvement in FPR95 over strong baselines while maintaining competitive ID accuracy.
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。