Local Augmentation for Graph Neural Networks

Data augmentation has been widely used in image data and linguistic data but remains under-explored for Graph Neural Networks (GNNs). Existing methods focus on augmenting the graph data from a global perspective and largely fall into two genres: structural manipulation and adversarial training with feature noise injection. However, recent graph data augmentation methods ignore the importance of local information for the GNNs' message passing mechanism. In this work, we introduce the local augmentation, which enhances the locality of node representations by their subgraph structures. Specifically, we model the data augmentation as a feature generation process. Given a node's features, our local augmentation approach learns the conditional distribution of its neighbors' features and generates more neighbors' features to boost the performance of downstream tasks. Based on the local augmentation, we further design a novel framework: LA-GNN, which can apply to any GNN models in a plug-and-play manner. Extensive experiments and analyses show that local augmentation consistently yields performance improvement for various GNN architectures across a diverse set of benchmarks.