Abstract:Graph anomaly detection (GAD), which aims to identify unusual graph instances (nodes, edges, subgraphs, or graphs), has attracted increasing attention in recent years due to its significance in a wide range of applications. Deep learning approaches, graph neural networks (GNNs) in particular, have been emerging as a promising paradigm for GAD, owing to its strong capability in capturing complex structure and/or node attributes in graph data. Considering the large number of methods proposed for GNN-based GAD, it is of paramount importance to summarize the methodologies and findings in the existing GAD studies, so that we can pinpoint effective model designs for tackling open GAD problems. To this end, in this work we aim to present a comprehensive review of deep learning approaches for GAD. Existing GAD surveys are focused on task-specific discussions, making it difficult to understand the technical insights of existing methods and their limitations in addressing some unique challenges in GAD. To fill this gap, we first discuss the problem complexities and their resulting challenges in GAD, and then provide a systematic review of current deep GAD methods from three novel perspectives of methodology, including GNN backbone design, proxy task design for GAD, and graph anomaly measures. To deepen the discussions, we further propose a taxonomy of 13 fine-grained method categories under these three perspectives to provide more in-depth insights into the model designs and their capabilities. To facilitate the experiments and validation, we also summarize a collection of widely-used GAD datasets and empirical comparison. We further discuss multiple open problems to inspire more future high-quality research. A continuously updated repository for datasets, links to the codes of algorithms, and empirical comparison is available at <a class="link-external link-https" href="https://github.com/mala-lab/Awesome-Deep-Graph-Anomaly-Detection" rel="external noopener nofollow">this https URL</a>.

Parallel Algorithms for Anomalous Subgraph Detection.

ASA: Adversary Situation Awareness Via Heterogeneous Graph Convolutional Networks.

A Nonparametric Approach to Uncovering Connected Anomalies by Tree Shaped Priors

Subgraph Centralization: A Necessary Step for Graph Anomaly Detection.

Efficient Nonparametric Subgraph Detection Using Tree Shaped Priors

A Universal Adaptive Algorithm for Graph Anomaly Detection

SubAnom: Efficient Subgraph Anomaly Detection Framework over Dynamic Graphs

Multi-representations Space Separation based Graph-level Anomaly-aware Detection

Near-optimal Anomaly Detection in Graphs using Lovasz Extended Scan Statistic

A Comprehensive Survey on Graph Anomaly Detection with Deep Learning

Selective ensemble method for anomaly detection based on parallel learning

AAAN: Anomaly Alignment in Attributed Networks

PSCAN: A Parallel Structural Clustering Algorithm for networks

Counterfactual Graph Learning for Anomaly Detection on Attributed Networks

Multitask Active Learning for Graph Anomaly Detection

MIRACLE: A Multiple Independent Random Walks Community Parallel Detection Algorithm for Big Graphs

Anomaly detection of unstructured big data via semantic analysis and dynamic knowledge graph construction

Sketch-Based Anomaly Detection in Streaming Graphs

CSCAD: Correlation Structure-Based Collective Anomaly Detection in Complex System

From Unsupervised to Few-shot Graph Anomaly Detection: A Multi-scale Contrastive Learning Approach

Deep Graph Anomaly Detection: A Survey and New Perspectives