Matcha: Mitigating Graph Structure Shifts with Test-Time Adaptation

2024-10-09

AI Summary

Graph Neural Networks (GNNs) suffer from representation degradation and blurred class boundaries during testing under graph structural shift—i.e., changes in node connectivity—rendering existing test-time adaptation (TTA) methods, designed for visual attribute shifts, ineffective. Method: We propose the first source-free, online TTA framework tailored specifically for structural shift. It dynamically adjusts message-passing parameters in GNNs, refines pseudo-labels via confidence-guided clustering, and employs lightweight gradient updates. A novel prediction-guided clustering loss is introduced to enhance representation discriminability. The framework is modular and compatible with mainstream TTA methods to jointly handle structural and attribute shifts. Contribution/Results: Evaluated on diverse synthetic and real-world graph benchmarks, our method achieves an average accuracy gain of 12.7% under structural-shift-dominant scenarios, while maintaining high computational efficiency and strong generalization robustness.

Technology Category

• Machine Learning: Graph-based Machine Learning
• Computer Vision: Adversarial Attacks & Robustness
• Reasoning under Uncertainty: Graphical Models

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Powerful as they are, graph neural networks (GNNs) are known to be vulnerable to distribution shifts. Recently, test-time adaptation (TTA) has attracted attention due to its ability to adapt a pre-trained model to a target domain, without re-accessing the source domain. However, existing TTA algorithms are primarily designed for attribute shifts in vision tasks, where samples are independent. These methods perform poorly on graph data that experience structure shifts, where node connectivity differs between source and target graphs. We attribute this performance gap to the distinct impact of node attribute shifts versus graph structure shifts: the latter significantly degrades the quality of node representations and blurs the boundaries between different node categories. To address structure shifts in graphs, we propose Matcha, an innovative framework designed for effective and efficient adaptation to structure shifts by adjusting the htop-aggregation parameters in GNNs. To enhance the representation quality, we design a prediction-informed clustering loss to encourage the formation of distinct clusters for different node categories. Additionally, Matcha seamlessly integrates with existing TTA algorithms, allowing it to handle attribute shifts effectively while improving overall performance under combined structure and attribute shifts. We validate the effectiveness of Matcha on both synthetic and real-world datasets, demonstrating its robustness across various combinations of structure and attribute shifts. Our code is available at https://github.com/baowenxuan/Matcha .

Problem

• Addresses graph neural networks' vulnerability to distribution shifts
• Focuses on mitigating structure shifts in graph data
• Proposes Matcha for efficient test-time adaptation in GNNs

Innovation

• Adapts GNNs to structure shifts
• Uses prediction-informed clustering loss
• Integrates with existing TTA algorithms