Graph Anomaly Detection via Multi-Scale Contrastive Learning Networks with Augmented View

Abstract

Graph anomaly detection (GAD) is a vital task in graph-based machine learning and has been widely applied in many real-world applications. The primary goal of GAD is to capture anomalous nodes from graph datasets, which evidently deviate from the majority of nodes. Recent methods have paid attention to various scales of contrastive strategies for GAD, i.e., node-subgraph and node-node contrasts. However, they neglect the subgraph-subgraph comparison information which the normal and abnormal subgraph pairs behave differently in terms of embeddings and structures in GAD, resulting in sub-optimal task performance. In this paper, we fulfill the above idea in the proposed multi-view multi-scale contrastive learning framework with subgraph-subgraph contrast for the first practice. To be specific, we regard the original input graph as the first view and generate the second view by graph augmentation with edge modifications. With the guidance of maximizing the similarity of the subgraph pairs, the proposed subgraph-subgraph contrast contributes to more robust subgraph embeddings despite of the structure variation. Moreover, the introduced subgraph-subgraph contrast cooperates well with the widely-adopted node-subgraph and node-node contrastive counterparts for mutual GAD performance promotions. Besides, we also conduct sufficient experiments to investigate the impact of different graph augmentation approaches on detection performance. The comprehensive experimental results well demonstrate the superiority of our method compared with the state-of-the-art approaches and the effectiveness of the multi-view subgraph pair contrastive strategy for the GAD task. The source code is released at https://github.com/FelixDJC/GRADATE.

ADA-GAD: Anomaly-Denoised Autoencoders for Graph Anomaly Detection

“…Anomaly Detection on Static Attributed Graphs Graph anomaly detection (Duan et al 2023) aims to identify nodes that are different from most nodes. Some progress has been made in anomaly detection on static attributed graphs.…”

“…The goal of unsupervised graph anomaly detection (GAD) is to identify rare patterns that deviate from the majority patterns in a graph, which has been extensively applied in diverse domains, such as fraud detection (Abdallah, Maarof, and Zainal 2016;Cheng et al 2020;Dou et al 2020) and social network (Fan, Zhang, and Li 2020;Duan et al 2023). Recently, reconstruction-based Graph Neural Networks (GNNs) methods have achieved great success and have become the mainstream approach.…”

GNAR: graph contrastive learning networks with adaptive readouts for anomaly detection

“…CoLA introduce the contrast-learning paradigm into GAD tasks has further advanced the field. Researchers have made additional improvements based on CoLA [17][18][19].…”

A graph encoder–decoder network for unsupervised anomaly detection

“…Zhuang et al [23] propose a subgraph centralization approach for graph anomaly detection, addressing the weaknesses of existing detectors in terms of computational cost, suboptimal detection accuracy, and lack of explanation for identified anomalies, leading to the development of a graph-centric anomaly detection framework. Duan et al [24] present a multi-view, multi-scale contrastive learning framework with subgraph-subgraph contrast for graph anomaly detection by combining various anomalous information and calculating the anomaly score for each node. However, most of these approaches do not incorporate pooling operations explicitly.…”