Backdoor Attacks on Graph Classification via Data Augmentation and Dynamic Poisoning

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

Graph neural networks (GNNs) have gained widespread adoption in domains such as bioinformatics, social networks, and cheminformatics, yet they remain susceptible to backdoor attacks. Existing backdoor attacks typically rely on subgraph triggers, which often introduce detectable anomalies and employ random poisoned sample selection, resulting in reduced stealthiness and efficiency. To address these limitations, we propose a novel backdoor attack framework that leverages data augmentation-based triggers and dynamic poisoned sample selection. Specifically, we design three alternative data augmentation strategies, edge modification guided by cosine similarity, edge removal based on degree centrality, and feature masking via gradient saliency, as backdoor triggers. Furthermore, we introduce a dynamic poisoned sample selection method informed by forgetting events. This method dynamically prioritizes high-impact poisoned samples to enhance attack efficiency while reducing the number of samples required to achieve the corresponding attack success rate (ASR). Experiments on four benchmark datasets, PROTEINS, NCI1, Mutagenicity, and ENZYMES, demonstrate the superiority of our method.

Original languageEnglish
Title of host publicationMachine Learning and Knowledge Discovery in Databases. Research Track - European Conference, ECML PKDD 2025, Proceedings
EditorsRita P. Ribeiro, Alípio M. Jorge, Bernhard Pfahringer, Nathalie Japkowicz, Pedro Larrañaga, Carlos Soares, Pedro H. Abreu, João Gama
PublisherSpringer Science and Business Media Deutschland GmbH
Pages263-279
Number of pages17
ISBN (Print)9783032060655
DOIs
Publication statusPublished - 2026
EventEuropean Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases, ECML PKDD 2025 - Porto, Portugal
Duration: 15 Sept 202519 Sept 2025

Publication series

NameLecture Notes in Computer Science
Volume16015 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

ConferenceEuropean Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases, ECML PKDD 2025
Country/TerritoryPortugal
CityPorto
Period15/09/2519/09/25

Keywords

  • Backdoor Attacks
  • GNNs
  • Graph Classification

Fingerprint

Dive into the research topics of 'Backdoor Attacks on Graph Classification via Data Augmentation and Dynamic Poisoning'. Together they form a unique fingerprint.

Cite this