mlsanityData to 7 Oct 2026

Defending Against Backdoor Attack on Graph Nerual Network by Explainability

arXiv 2209.02902 ↗First posted 7 Sep 2022cs.AIcited by 9

Not on the map. Why?

Abstract (arXiv)

Backdoor attack is a powerful attack algorithm to deep learning model. Recently, GNN's vulnerability to backdoor attack has been proved especially on graph classification task. In this paper, we propose the first backdoor detection and defense method on GNN. Most backdoor attack depends on injecting small but influential trigger to the clean sample. For graph data, current backdoor attack focus on manipulating the graph structure to inject the trigger. We find that there are apparent differences between benign samples and malicious samples in some explanatory evaluation metrics, such as fidelity and infidelity. After identifying the malicious sample, the explainability of the GNN model can help us capture the most significant subgraph which is probably the trigger in a trojan graph. We use various dataset and different attack settings to prove the effectiveness of our defense method. The attack success rate all turns out to decrease considerably.

Key facts. Cited by 9 papers in the map (data through 7 Oct 2026). Builds on 29 of its 43 references that are in the map. First posted on arXiv 7 Sep 2022.

Builds on

29 of 43 references.

References · 29

  1. 2014Explaining and Harnessing Adversarial Examplescited by 7,287
  2. 2018How Powerful are Graph Neural Networks?cited by 4,355
  3. 2017Axiomatic Attribution for Deep Networkscited by 3,350
  4. 2019Fast Graph Representation Learning with PyTorch Geometriccited by 2,414
  5. 2018Obfuscated Gradients Give a False Sense of Security: Circumventing Defenses to Adversarial Examplescited by 1,259
  6. 2017BadNets: Identifying Vulnerabilities in the Machine Learning Model Supply Chaincited by 1,052
  7. 2017Targeted Backdoor Attacks on Deep Learning Systems Using Data Poisoningcited by 1,012
  8. 2018Weisfeiler and Leman Go Neural: Higher-order Graph Neural Networkscited by 878
  9. 2019GNNExplainer: Generating Explanations for Graph Neural Networkscited by 735
  10. 2018A Benchmark for Interpretability Methods in Deep Neural Networkscited by 467
  11. 2018Spectral Signatures in Backdoor Attackscited by 457
  12. 2019STRIP: A Defence Against Trojan Attacks on Deep Neural Networkscited by 437
17 more
  1. 2020Parameterized Explainer for Graph Neural Networkcited by 423
  2. 2018Adversarial Attacks on Neural Networks for Graph Datacited by 395
  3. 2018Adversarial Attack on Graph Structured Datacited by 300
  4. 2017Detecting Adversarial Samples from Artifactscited by 244
  5. 2020XGNN: Towards Model-Level Explanations of Graph Neural Networkscited by 211
  6. 2020PGM-Explainer: Probabilistic Graphical Model Explanations for Graph Neural Networkscited by 210
  7. 2020GraphLIME: Local Interpretable Model Explanations for Graph Neural Networkscited by 209
  8. 2020Interpreting Graph Neural Networks for NLP With Differentiable Edge Maskingcited by 152
  9. 2020Higher-Order Explanations of Graph Neural Networks via Relevant Walkscited by 138
  10. 2020Backdoor Attacks to Graph Neural Networkscited by 126
  11. 2021ML-Doctor: Holistic Risk Assessment of Inference Attacks Against Machine Learning Modelscited by 79
  12. 2017Generating Adversarial Malware Examples for Black-Box Attacks Based on GANcited by 74
  13. 2017Efficient Defenses Against Adversarial Attackscited by 53
  14. 2020RelEx: A Model-Agnostic Relational Model Explainercited by 53
  15. 2021Explainability-based Backdoor Attacks Against Graph Neural Networkscited by 42
  16. 2019Layerwise Relevance Visualization in Convolutional Text Graph Classifierscited by 32
  17. 2018Hu-Fu: Hardware and Software Collaborative Attack Framework against Neural Networkscited by 8

Built on by

Cited by year

Citing papers per year: 2023: 3; 2024: 5; 2025: 1.

9 papers.

  1. 2023A clean-label graph backdoor attack method in node classification taskcited by 10
  2. 2023XGBD: Explanation-Guided Graph Backdoor Detectioncited by 6
  3. 2024Identifying Backdoored Graphs in Graph Neural Network Training: An Explanation-Based Approach with Novel Metricscited by 6
  4. 2024Graph Neural Backdoor: Fundamentals, Methodologies, Applications, and Future Directionscited by 4
  5. 2023MDTD: A Multi Domain Trojan Detector for Deep Neural Networkscited by 1
  6. 2024MADE: Graph Backdoor Defense with Masked Unlearningcited by 1
  7. 2025A Semantic and Clean-label Backdoor Attack against Graph Convolutional Networkscited by 1
  8. 2024"No Matter What You Do": Purifying GNN Models via Backdoor Unlearningcited by 0
  9. 2024DMGNN: Detecting and Mitigating Backdoor Attacks in Graph Neural Networkscited by 0