Balanced Anomaly-guided Ego-graph Diffusion Model for Inductive Graph Anomaly Detection

TL;DR

This paper introduces a novel graph anomaly detection framework that combines dynamic ego-graph diffusion modeling with balanced synthetic anomaly data augmentation to improve detection in evolving networks.

Contribution

It proposes a data-centric approach integrating ego-graph diffusion and curriculum-based anomaly synthesis for inductive graph anomaly detection.

Findings

Effective in detecting anomalies across five datasets.

Improves generalization in dynamic, imbalanced graph scenarios.

Outperforms existing methods in experimental evaluations.

Abstract

Graph anomaly detection (GAD) is crucial in applications like fraud detection and cybersecurity. Despite recent advancements using graph neural networks (GNNs), two major challenges persist. At the model level, most methods adopt a transductive learning paradigm, which assumes static graph structures, making them unsuitable for dynamic, evolving networks. At the data level, the extreme class imbalance, where anomalous nodes are rare, leads to biased models that fail to generalize to unseen anomalies. These challenges are interdependent: static transductive frameworks limit effective data augmentation, while imbalance exacerbates model distortion in inductive learning settings. To address these challenges, we propose a novel data-centric framework that integrates dynamic graph modeling with balanced anomaly synthesis. Our framework features: (1) a discrete ego-graph diffusion model, which captures the local topology of anomalies to generate ego-graphs aligned with anomalous structural distribution, and (2) a curriculum anomaly augmentation mechanism, which dynamically adjusts synthetic data generation during training, focusing on underrepresented anomaly patterns to improve detection and generalization. Experiments on five datasets demonstrate that the effectiveness of our framework.