Rank and Align: Towards Effective Source-free Graph Domain Adaptation

TL;DR

This paper introduces Rank and Align (RNA), a novel source-free GNN approach that uses spectral seriation and harmonic graph alignment to adapt to new domains without source data, addressing privacy concerns.

Contribution

RNA is the first to combine spectral seriation and harmonic graph alignment for source-free graph domain adaptation, enabling effective transfer without source graphs.

Findings

RNA outperforms existing methods on benchmark datasets.

Spectral seriation improves semantic learning under domain shift.

Adversarial subgraph extraction reduces domain discrepancy.

Abstract

Graph neural networks (GNNs) have achieved impressive performance in graph domain adaptation. However, extensive source graphs could be unavailable in real-world scenarios due to privacy and storage concerns. To this end, we investigate an underexplored yet practical problem of source-free graph domain adaptation, which transfers knowledge from source models instead of source graphs to a target domain. To solve this problem, we introduce a novel GNN-based approach called Rank and Align (RNA), which ranks graph similarities with spectral seriation for robust semantics learning, and aligns inharmonic graphs with harmonic graphs which close to the source domain for subgraph extraction. In particular, to overcome label scarcity, we employ the spectral seriation algorithm to infer the robust pairwise rankings, which can guide semantic learning using a similarity learning objective. To depict distribution shifts, we utilize spectral clustering and the silhouette coefficient to detect harmonic graphs, which the source model can easily classify. To reduce potential domain discrepancy, we extract domain-invariant subgraphs from inharmonic graphs by an adversarial edge sampling process, which guides the invariant learning of GNNs. Extensive experiments on several benchmark datasets demonstrate the effectiveness of our proposed RNA.