Scaling Graph Neural Networks with Approximate PageRank

TL;DR

The paper introduces PPRGo, a scalable and efficient GNN model that approximates information diffusion using PageRank, enabling fast training and prediction on large graphs with minimal computational resources.

Contribution

PPRGo leverages an approximate PageRank method to significantly improve scalability and speed of GNNs without sacrificing prediction accuracy.

Findings

PPRGo outperforms baselines in academic graph tasks.

Training on a large graph with 12.4 million nodes takes under 2 minutes.

PPRGo is easily parallelizable and suitable for industry-scale applications.

Abstract

Graph neural networks (GNNs) have emerged as a powerful approach for solving many network mining tasks. However, learning on large graphs remains a challenge - many recently proposed scalable GNN approaches rely on an expensive message-passing procedure to propagate information through the graph. We present the PPRGo model which utilizes an efficient approximation of information diffusion in GNNs resulting in significant speed gains while maintaining state-of-the-art prediction performance. In addition to being faster, PPRGo is inherently scalable, and can be trivially parallelized for large datasets like those found in industry settings. We demonstrate that PPRGo outperforms baselines in both distributed and single-machine training environments on a number of commonly used academic graphs. To better analyze the scalability of large-scale graph learning methods, we introduce a novel benchmark graph with 12.4 million nodes, 173 million edges, and 2.8 million node features. We show that training PPRGo from scratch and predicting labels for all nodes in this graph takes under 2 minutes on a single machine, far outpacing other baselines on the same graph. We discuss the practical application of PPRGo to solve large-scale node classification problems at Google.