USBD: Universal Structural Basis Distillation for Source-Free Graph Domain Adaptation

2026-02-09

AI Summary

This work addresses the performance degradation in source-agnostic graph domain adaptation caused by structural biases inherent in source models, particularly when target graphs exhibit substantial topological discrepancies. To overcome this limitation, the authors propose a novel paradigm that constructs a universal, structure-agnostic basis. This is achieved through bilevel optimization to distill a set of prototypes from the source graph that collectively span the full spectrum of topological patterns. During inference, the method dynamically activates an optimal combination of these prototypes based on the spectral characteristics of the target graph. By explicitly modeling the complete Dirichlet energy spectrum, the approach transcends the constraints of conventional pseudo-labeling methods under drastic structural shifts. Extensive experiments demonstrate significant performance gains over state-of-the-art methods across multiple benchmarks, especially in scenarios with large structural divergence, while maintaining computational efficiency decoupled from the scale of target data.

Technology Category

• Machine Learning: Graph-based Machine Learning
• Search and Optimization: Sampling/Simulation-based Search
• Constraint Satisfaction and Optimization: Distributed CSP/Optimization

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Semantics and Knowledge: Scalable techniques for the creation, curation, publication, maintenance, and consumption of large, Web-based, structured, reusable, knowledge graphs and ontologies
• Search and Retrieval-Augmented AI: Vertical and domain-specific search

Abstract

SF-GDA is pivotal for privacy-preserving knowledge transfer across graph datasets. Although recent works incorporate structural information, they implicitly condition adaptation on the smoothness priors of sourcetrained GNNs, thereby limiting their generalization to structurally distinct targets. This dependency becomes a critical bottleneck under significant topological shifts, where the source model misinterprets distinct topological patterns unseen in the source domain as noise, rendering pseudo-label-based adaptation unreliable. To overcome this limitation, we propose the Universal Structural Basis Distillation, a framework that shifts the paradigm from adapting a biased model to learning a universal structural basis for SF-GDA. Instead of adapting a biased source model to a specific target, our core idea is to construct a structure-agnostic basis that proactively covers the full spectrum of potential topological patterns. Specifically, USBD employs a bi-level optimization framework to distill the source dataset into a compact structural basis. By enforcing the prototypes to span the full Dirichlet energy spectrum, the learned basis explicitly captures diverse topological motifs, ranging from low-frequency clusters to high-frequency chains, beyond those present in the source. This ensures that the learned basis creates a comprehensive structural covering capable of handling targets with disparate structures. For inference, we introduce a spectral-aware ensemble mechanism that dynamically activates the optimal prototype combination based on the spectral fingerprint of the target graph. Extensive experiments on benchmarks demonstrate that USBD significantly outperforms state-of-the-art methods, particularly in scenarios with severe structural shifts, while achieving superior computational efficiency by decoupling the adaptation cost from the target data scale.

Problem

• Source-Free Graph Domain Adaptation
• Structural Shift
• Topological Patterns
• Graph Neural Networks
• Domain Adaptation

Innovation

• Universal Structural Basis
• Source-Free Graph Domain Adaptation
• Dirichlet Energy Spectrum
• Spectral-Aware Ensemble
• Topology-Agnostic Representation