Graph Transplant: Node Saliency-Guided Graph Mixup with Local Structure Preservation

TL;DR

Graph Transplant introduces a novel graph augmentation technique that mixes subgraphs based on node saliency, improving graph classification performance, robustness, and calibration across diverse datasets.

Contribution

This paper presents the first Mixup-like graph augmentation method that preserves local structure and employs node saliency for meaningful subgraph selection.

Findings

Outperforms basic data augmentation baselines in accuracy.

Enhances robustness and model calibration.

Effective across diverse graph datasets and architectures.

Abstract

Graph-structured datasets usually have irregular graph sizes and connectivities, rendering the use of recent data augmentation techniques, such as Mixup, difficult. To tackle this challenge, we present the first Mixup-like graph augmentation method at the graph-level called Graph Transplant, which mixes irregular graphs in data space. To be well defined on various scales of the graph, our method identifies the sub-structure as a mix unit that can preserve the local information. Since the mixup-based methods without special consideration of the context are prone to generate noisy samples, our method explicitly employs the node saliency information to select meaningful subgraphs and adaptively determine the labels. We extensively validate our method with diverse GNN architectures on multiple graph classification benchmark datasets from a wide range of graph domains of different sizes. Experimental results show the consistent superiority of our method over other basic data augmentation baselines. We also demonstrate that Graph Transplant enhances the performance in terms of robustness and model calibration.