Connecting the dots: Exploring backdoor attacks on graph neural networks

Computer Science

• Backdoors 100%
• Attack 100%
• Graph Neural Network 100%
• Deep Neural Network 21%
• Model 10%
• Neural Network Model 10%
• Federated Graph 10%
• Real World 10%
• Domain 10%
• Design 5%
• Attackers 5%
• Classifier 5%
• Watermarking 5%
• Application 5%
• Privacy Concern 5%
• Speech Recognition 5%
• Adversarial Machine Learning 5%
• Natural Language Processing 5%
• Image Classification 5%

INIS

• graphs 100%
• neural networks 100%
• data 17%
• images 13%
• processing 8%
• performance 8%
• increasing 4%
• design 4%
• exploration 4%
• applications 4%
• datasets 4%
• efficiency 4%
• speech 4%
• natural language 4%
• interactions 4%
• cars 4%
• output 4%
• security 4%
• speed limit 4%
• classification 4%