GNN Transformation Framework for Improving Efficiency and Scalability

TL;DR

This paper introduces a framework that automatically transforms non-scalable GNNs into efficient, precomputation-based models suitable for large-scale graphs, significantly improving training speed while maintaining competitive accuracy.

Contribution

The framework enables scalable GNNs by separating local aggregation from weight learning and efficiently executing precomputation on GPUs, providing a new approach for large-scale graph processing.

Findings

Transformed GNNs run faster in training time than existing GNNs.

Achieves competitive accuracy with state-of-the-art GNNs on large-scale graphs.

Provides simple, efficient baselines for scalable GNN research.

Abstract

We propose a framework that automatically transforms non-scalable GNNs into precomputation-based GNNs which are efficient and scalable for large-scale graphs. The advantages of our framework are two-fold; 1) it transforms various non-scalable GNNs to scale well to large-scale graphs by separating local feature aggregation from weight learning in their graph convolution, 2) it efficiently executes precomputation on GPU for large-scale graphs by decomposing their edges into small disjoint and balanced sets. Through extensive experiments with large-scale graphs, we demonstrate that the transformed GNNs run faster in training time than existing GNNs while achieving competitive accuracy to the state-of-the-art GNNs. Consequently, our transformation framework provides simple and efficient baselines for future research on scalable GNNs.