Improving Graph Neural Networks at Scale: Combining Approximate PageRank and CoreRank
This work addresses scalability problems for researchers and practitioners using GNNs on industrial-scale graphs, representing an incremental improvement over existing methods.
The paper tackles the scalability issue of Graph Neural Networks (GNNs) on large graphs by proposing CorePPR, a model that combines approximate PageRank and CoreRank to efficiently diffuse information, and it outperforms the baseline PPRGo, especially on large graphs.
Graph Neural Networks (GNNs) have achieved great successes in many learning tasks performed on graph structures. Nonetheless, to propagate information GNNs rely on a message passing scheme which can become prohibitively expensive when working with industrial-scale graphs. Inspired by the PPRGo model, we propose the CorePPR model, a scalable solution that utilises a learnable convex combination of the approximate personalised PageRank and the CoreRank to diffuse multi-hop neighbourhood information in GNNs. Additionally, we incorporate a dynamic mechanism to select the most influential neighbours for a particular node which reduces training time while preserving the performance of the model. Overall, we demonstrate that CorePPR outperforms PPRGo, particularly on large graphs where selecting the most influential nodes is particularly relevant for scalability. Our code is publicly available at: https://github.com/arielramos97/CorePPR.