Improving Attention Mechanism in Graph Neural Networks via Cardinality Preservation

4 Jul 2019  ·  Shuo Zhang, Lei Xie ·

Graph Neural Networks (GNNs) are powerful to learn the representation of graph-structured data. Most of the GNNs use the message-passing scheme, where the embedding of a node is iteratively updated by aggregating the information of its neighbors. To achieve a better expressive capability of node influences, attention mechanism has grown to be popular to assign trainable weights to the nodes in aggregation. Though the attention-based GNNs have achieved remarkable results in various tasks, a clear understanding of their discriminative capacities is missing. In this work, we present a theoretical analysis of the representational properties of the GNN that adopts the attention mechanism as an aggregator. Our analysis determines all cases when those attention-based GNNs can always fail to distinguish certain distinct structures. Those cases appear due to the ignorance of cardinality information in attention-based aggregation. To improve the performance of attention-based GNNs, we propose cardinality preserved attention (CPA) models that can be applied to any kind of attention mechanisms. Our experiments on node and graph classification confirm our theoretical analysis and show the competitive performance of our CPA models.

PDF Abstract

Results from the Paper


Task Dataset Model Metric Name Metric Value Global Rank Result Benchmark
Graph Classification ENZYMES GAT-GC (f-Scaled) Accuracy 58.45 # 24
Graph Classification MUTAG GAT-GC (f-Scaled) Accuracy 90.44% # 20
Graph Classification NCI1 GAT-GC (f-Scaled) Accuracy 82.28% # 24
Graph Classification PROTEINS GAT-GC (f-Scaled) Accuracy 76.81% # 32
Graph Classification REDDIT-B GAT-GC (f-Scaled) Accuracy 92.57 # 2
Graph Classification RE-M5K GAT-GC (f-Scaled) Accuracy 57.22% # 2

Methods