Deep Graph Anomaly Detection: A Survey and New Perspectives

A professionally curated list of awesome resources (paper, code, data, etc.) on Deep Graph Anomaly Detection (DGAD), which is the first work to comprehensively and systematically summarize the recent advances of deep graph anomaly detection from the methodology design to the best of our knowledge.

We will continue to update this list with the latest resources. If you find any missed resources (paper/code) or errors, please feel free to open an issue or make a pull request.

Tutorial

📢 We are delighted to share that we have successfully hosted the <strong>tutorial</strong> titled 👉 <a href="https://sites.google.com/view/ijcai-tutorial-on-ad/home"><strong>Deep Learning for Graph Anomaly Detection</strong></a> at the <a href="https://2025.ijcai.org/"><strong>IJCAI 2025</strong></a>!

This tutorial is based on our recent comprehensive <strong>survey</strong> in the field.

The tutorial slides can be obtained here <a href="https://drive.google.com/file/d/17qbeiLvx9r8bf-cmrO4vhKBYSHaRcZWO/view?usp=drive_link"><strong>Slides</strong></a>

Survey Paper (IEEE TKDE 2025)

Deep Graph Anomaly Detection: A Survey and New Perspectives

Hezhe Qiao, Hanghang Tong, Bo An, Irwin King, Charu Aggarwal, Guansong Pang.

If you find this repository helpful for your work, please kindly cite our paper.

@article{qiao2025deep,
  title={Deep Graph Anomaly Detection: A Survey and New Perspectives},
  author={Qiao, Hezhe and Tong, Hanghang and An, Bo and King, Irwin and Aggarwal, Charu and Pang, Guansong},
  journal={IEEE Transactions on Knowledge and Data Engineering},
  year={2025},
  publisher={IEEE}
}

Feel free to point out any mistakes and welcome to provide relevant papers.

Taxonomy of Deep Graph Anomaly Detection

<img src="overview.png" width=900 align=middle> <br />

Outline

The outline corresponds to the taxonomy of methods in our survey paper.

• 1. GNN Backbone Design
  • 1.1 Discriminative GNNs
    • 1.1.1 Aggregation Mechanism
    • 1.1.2 Feature Transformation
  • 1.2 Generative GNNs
    • 1.2.1 Feature Interpolation
    • 1.2.2 Noise Perturbation
• 2. Proxy Task Design
  • 2.1 Graph Reconstruction
  • 2.2 Graph Contrastive Learning
  • 2.3 Graph Representation Distillati
  • 2.4 Adversarial Graph Learning
  • 2.5 Score Prediction
• 3. Graph Anomaly Measures
  • 3.1 One-class Classification Measure
  • 3.2 Community Adherence
  • 3.3 Local Affinity
  • 3.4 Graph Isolation
• 4. Graph Anomaly Detection Related Survey
• 5. Anomaly Detection Related Survey
• 6. Quantitative Comparison
• 7. Datasets

Categories of Deep Graph Anomaly Detection

GNN Backbone Design

<img src="Fig1.png" width=900 align=middle> <br />

Discriminative GNNs

Aggregation Mechanism

• [Dou2020] Enhancing Graph Neural Network-based Fraud Detectors against Camouflaged Fraudsters in CIKM, 2020. [paper][code]

• [Liu2020] Alleviating the Inconsistency Problem of Applying Graph Neural Network to Fraud Detection in SIGIR, 2020. [paper][code]

• [Liu2021] Pick and Choose: A GNN-based Imbalanced Learning Approach for Fraud Detection in WWW, 2021.[paper][code]

• [Zhang2021] FRAUDRE: Fraud Detection Dual-Resistant to Graph Inconsistency and Imbalance in ICDM, 2021. [paper][code]

• [Zhang2022] Dual-discriminative Graph Neural Network for Imbalanced Graph-level Anomaly Detection in NeurIPS, 2022. [paper]

• [Qin2022] Explainable Graph-based Fraud Detection via Neural Meta-graph Search in CIKM, 2022. [paper][code]

• [Dong2022] Bi-Level Selection via Meta Gradient for Graph-based Fraud Detection in DASFAA, 2022. [paper]

• [Shi2022] H2-FDetector: A GNN-based Fraud Detector with Homophilic and Heterophilic Connections in WebConf, 2022. [paper]

• [Gao2023] Addressing Heterophily in Graph Anomaly Detection: A Perspective of Graph Spectrum in WebConf, 2023. [paper][code]

• [Ma2023] Towards Graph-level Anomaly Detection via Deep Evolutionary Mapping in KDD, 2023. [paper][code]

• [Chang2024] Multitask Active Learning for Graph Anomaly Detection in Arxiv, 2024. [paper][code]

• [Zhang2024] Generation is better than Modification: Combating High Class Homophily Variance in Graph Anomaly Detection in Arxiv, 2024. [paper]

• [Chen2024] Boosting Graph Anomaly Detection with Adaptive Message Passing in ICLR, 2024. [paper]

• [Zhuo2023] Partitioning Message Passing for Graph Fraud Detection in ICLR, 2024. [paper][code]

• [Gao2024] Graph Anomaly Detection with Bi-level Optimization in WebConf, 2024. [paper][code]

• [Guo2024] Graph Local Homophily Network for Anomaly Detection in CIKM, 2024. [paper]

• [Ma2024] Revisiting Score Propagation in Graph Out-of-Distribution Detection in NeurIPS, 2024. [paper][code]

• [Hyun2024] LEX-GNN: Label-Exploring Graph Neural Network for Accurate Fraud Detection in CIKM, 2024. [paper][code]

• [Liu2024] Collaborative Fraud Detection on Large Scale Graph Using Secure Multi-Party Computation in CIKM, 2024. [paper]

• [Yang2024] A Generalizable Anomaly Detection Method in Dynamic Graph in AAAI, 2024. [paper][code]

• [Haghighi] TROPICAL: Transformer-based Hypergraph Learning for Camouflaged Fraudster Detection in ICDM, 2024. [paper][code]

• [Wang2025] Open-set Graph Anomaly Detection via Normal Structure Regularisation in ICLR, 2025. [paper][code]

• [Hao2025] Bi-directional Curriculum Learning for Graph Anomaly Detection: Dual Focus on Homogeneity and Heterogeneity in Arxiv, 2025. [[pape