Scaling Graph Neural Networks with Approximate PageRank
This addresses scalability issues in GNNs for large-scale node classification problems, particularly in industry settings like Google, though it is incremental as it builds on existing GNN methods with a novel approximation technique.
The paper tackles the challenge of scaling graph neural networks (GNNs) for large graphs by introducing PPRGo, which uses approximate PageRank to speed up information diffusion, achieving state-of-the-art performance with training and prediction on a 12.4-million-node graph in under 2 minutes on a single machine.
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.