arXiv 2410.148862025-06-07cs.LG

基于统一邻域提示的零样本通用图异常检测

Zero-shot Generalist Graph Anomaly Detection with Unified Neighborhood Prompts

AAII, University of Technology Sydney（AAII，悉尼大学）；School of Computing and Information Systems, Singapore Management University（计算机与信息系统学院，新加坡管理大学）；Faculty of Data Science, City University of Macau（数据科学学院，澳门城市大学）

Chaoxi Niu, Hezhe Qiao, Changlu Chen, Ling Chen, Guansong Pang

备注：: Accepted by IJCAI25

阅读 PDF

arXiv 页面 ↗HTML 原文 ↗

AI总结：

本文提出UNPrompt零样本通用图异常检测方法，通过逐坐标归一化和可学习的通用邻域提示，将潜在节点属性可预测性作为跨数据集异常分数，实现单数据集训练后免微调泛化。

摘要：

图异常检测（GAD）旨在识别图中显著偏离正常模式的节点，在广泛的应用领域中发挥着关键作用。然而，现有的GAD方法采用“一个模型对应一个数据集”的方式，即为每个图数据集训练单独的模型，这在很大程度上限制了其在现实场景中的适用性。为克服这一局限，我们提出了一种新颖的零样本通用GAD方法UNPrompt，用于训练“一个模型适配全部”的检测模型：只需在单个图数据集上训练一个GAD模型，随后即可有效泛化到其他图数据集的异常检测，而无需任何重新训练或微调。UNPrompt的核心洞见在于：i）潜在节点属性的可预测性可作为一种通用异常度量；ii）在经过适当归一化的节点属性空间中，可通过潜在节点属性预测学习通用的正常与异常图模式。UNPrompt通过两个主要模块实现GAD的通用模式：一个模块通过逐坐标归一化对齐不同图之间节点属性的维度和语义，另一个模块学习通用邻域提示，以支持在不同数据集上将潜在节点属性的可预测性用作异常分数。在真实世界GAD数据集上的大量实验表明，UNPrompt在通用GAD设置下显著优于多种竞争方法，并且在“一个模型对应一个数据集”设置下也具有很强的优势。代码可在https://github.com/mala-lab/UNPrompt获取。

英文摘要：

Graph anomaly detection (GAD), which aims to identify nodes in a graph that significantly deviate from normal patterns, plays a crucial role in broad application domains. However, existing GAD methods are one-model-for-one-dataset approaches, i.e., training a separate model for each graph dataset. This largely limits their applicability in real-world scenarios. To overcome this limitation, we propose a novel zero-shot generalist GAD approach UNPrompt that trains a one-for-all detection model, requiring the training of one GAD model on a single graph dataset and then effectively generalizing to detect anomalies in other graph datasets without any retraining or fine-tuning. The key insight in UNPrompt is that i) the predictability of latent node attributes can serve as a generalized anomaly measure and ii) generalized normal and abnormal graph patterns can be learned via latent node attribute prediction in a properly normalized node attribute space. UNPrompt achieves a generalist mode for GAD through two main modules: one module aligns the dimensionality and semantics of node attributes across different graphs via coordinate-wise normalization, while another module learns generalized neighborhood prompts that support the use of latent node attribute predictability as an anomaly score across different datasets. Extensive experiments on real-world GAD datasets show that UNPrompt significantly outperforms diverse competing methods under the generalist GAD setting, and it also has strong superiority under the one-model-for-one-dataset setting. Code is available at https://github.com/mala-lab/UNPrompt.

机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。