Graph Neural Networks (GNN) is an emerging field for learning on non-Euclidean data. Recently, there has been increased interest in designing GNN that scales to large graphs. Most existing methods use "graph sampling" or "layer-wise sampling" techniques to reduce training time. However, these methods still suffer from degrading performance and scalability problems when applying to graphs with billions of edges. This paper presents GBP, a scalable GNN that utilizes a localized bidirectional propagation process from both the feature vectors and the training/testing nodes. Theoretical analysis shows that GBP is the first method that achieves sub-linear time complexity for both the precomputation and the training phases. An extensive empirical study demonstrates that GBP achieves state-of-the-art performance with significantly less training/testing time. Most notably, GBP can deliver superior performance on a graph with over 60 million nodes and 1.8 billion edges in less than half an hour on a single machine. The codes of GBP can be found at https://github.com/chennnM/GBP .
@article{arxiv.2010.15421,
title = {Scalable Graph Neural Networks via Bidirectional Propagation},
author = {Ming Chen and Zhewei Wei and Bolin Ding and Yaliang Li and Ye Yuan and Xiaoyong Du and Ji-Rong Wen},
journal= {arXiv preprint arXiv:2010.15421},
year = {2021}
}