Rethinking pooling in graph neural networks

TL;DR

This paper critically evaluates the importance of locality-preserving pooling in GNNs, showing that convolutional layers are more influential than pooling strategies for model performance.

Contribution

It introduces variants of GNN pooling that challenge the necessity of locality-preserving methods and demonstrates their effectiveness through extensive experiments.

Findings

Local pooling is not essential for GNN success.

Convolutional layers primarily determine learned representations.

Randomized or complement graph clustering pooling do not degrade performance.

Abstract

Graph pooling is a central component of a myriad of graph neural network (GNN) architectures. As an inheritance from traditional CNNs, most approaches formulate graph pooling as a cluster assignment problem, extending the idea of local patches in regular grids to graphs. Despite the wide adherence to this design choice, no work has rigorously evaluated its influence on the success of GNNs. In this paper, we build upon representative GNNs and introduce variants that challenge the need for locality-preserving representations, either using randomization or clustering on the complement graph. Strikingly, our experiments demonstrate that using these variants does not result in any decrease in performance. To understand this phenomenon, we study the interplay between convolutional layers and the subsequent pooling ones. We show that the convolutions play a leading role in the learned representations. In contrast to the common belief, local pooling is not responsible for the success of GNNs on relevant and widely-used benchmarks.