EdgeMask-DG*: Learning Domain-Invariant Graph Structures via Adversarial Edge Masking

TL;DR

This paper introduces EdgeMask-DG*, a novel adversarial masking approach that identifies domain-invariant graph structures by combining topology and feature similarity, leading to improved domain generalization in graph neural networks.

Contribution

EdgeMask-DG* systematically combines adaptive adversarial topology search with feature-enriched graphs for better domain-invariant learning.

Findings

Achieves state-of-the-art results on graph domain generalization benchmarks.

Improves worst-case domain accuracy on Cora OOD benchmark by 3.8 percentage points.

Demonstrates robustness of the method across diverse graph types.

Abstract

Structural shifts pose a significant challenge for graph neural networks, as graph topology acts as a covariate that can vary across domains. Existing domain generalization methods rely on fixed structural augmentations or training on globally perturbed graphs, mechanisms that do not pinpoint which specific edges encode domain-invariant information. We argue that domain-invariant structural information is not rigidly tied to a single topology but resides in the consensus across multiple graph structures derived from topology and feature similarity. To capture this, we first propose EdgeMask-DG, a novel min-max algorithm where an edge masker learns to find worst-case continuous masks subject to a sparsity constraint, compelling a task GNN to perform effectively under these adversarial structural perturbations. Building upon this, we introduce EdgeMask-DG*, an extension that applies this adversarial masking principle to an enriched graph. This enriched graph combines the original topology with feature-derived edges, allowing the model to discover invariances even when the original topology is noisy or domain-specific. EdgeMask-DG* is the first to systematically combine adaptive adversarial topology search with feature-enriched graphs. We provide a formal justification for our approach from a robust optimization perspective. We demonstrate that EdgeMask-DG* achieves new state-of-the-art performance on diverse graph domain generalization benchmarks, including citation networks, social networks, and temporal graphs. Notably, on the Cora OOD benchmark, EdgeMask-DG* lifts the worst-case domain accuracy to 78.0\%, a +3.8 pp improvement over the prior state of the art (74.2\%). The source code for our experiments can be found here: https://anonymous.4open.science/r/TMLR-EAEF/