A Causality-Aligned Structure Rationalization Scheme Against Adversarial Biased Perturbations for Graph Neural Networks-Reference-Cited by-同舟云学术

A Causality-Aligned Structure Rationalization Scheme Against Adversarial Biased Perturbations for Graph Neural Networks

Published:2024 Issue: Volume:19 Page:59-73
ISSN:1556-6013
Container-title:IEEE Transactions on Information Forensics and Security
language:
Short-container-title:IEEE Trans.Inform.Forensic Secur.

Author:

Jia Ju¹^ORCID,Ma Siqi²^ORCID,Liu Yang³^ORCID,Wang Lina⁴^ORCID,Deng Robert H.⁵^ORCID

Affiliation:

1. School of Cyber Science and Engineering, Southeast University, Nanjing, China

2. School of System and Computing, University of New South Wales, Canberra, ACT, Australia

3. School of Computer Science and Engineering, Nanyang Technological University, Jurong West, Singapore

4. Key Laboratory of Aerospace Information Security and Trusted Computing, Ministry of Education, School of Cyber Science and Engineering, Wuhan University, Wuhan, China

5. School of Information Systems, Singapore Management University, Bras Basah, Singapore

Funder

National Natural Science Foundation of China

Start-Up Research Fund of Southeast University

National Research Foundation, Singapore, and Defence Science Organisation (DSO) National Laboratories through the AI Singapore Program

Cyber Security Agency through the National Cybersecurity Research and Development Program

Publisher

Institute of Electrical and Electronics Engineers (IEEE)

Subject

Computer Networks and Communications,Safety, Risk, Reliability and Quality

Link

http://xplorestaging.ieee.org/ielx7/10206/10319981/10262126.pdf?arnumber=10262126

Reference58 articles.

1. Deep Neural Backdoor in Semi-Supervised Learning: Threats and Countermeasures

2. Structural entropy guided graph hierarchical pooling;wu;Proc ICML,2022

3. Text Backdoor Detection Using an Interpretable RNN Abstract Model

4. Graph attention networks;velickovic;Proc ICLR,2018

5. Recognizing Predictive Substructures with Subgraph Information Bottleneck