SignGT: Signed Attention-based Graph Transformer for Graph Representation Learning

TL;DR

SignGT introduces a signed attention mechanism and a structure-aware feed-forward network to enhance graph Transformers, enabling better capture of diverse frequency information and local topology for improved graph representation learning.

Contribution

The paper proposes SignGT, a novel graph Transformer with signed self-attention and neighborhood bias, addressing limitations in capturing high-frequency information and local structure.

Findings

Outperforms state-of-the-art graph Transformers on multiple tasks.

Effectively captures both low- and high-frequency information.

Improves learning of local topology and long-range dependencies.

Abstract

The emerging graph Transformers have achieved impressive performance for graph representation learning over graph neural networks (GNNs). In this work, we regard the self-attention mechanism, the core module of graph Transformers, as a two-step aggregation operation on a fully connected graph. Due to the property of generating positive attention values, the self-attention mechanism is equal to conducting a smooth operation on all nodes, preserving the low-frequency information. However, only capturing the low-frequency information is inefficient in learning complex relations of nodes on diverse graphs, such as heterophily graphs where the high-frequency information is crucial. To this end, we propose a Signed Attention-based Graph Transformer (SignGT) to adaptively capture various frequency information from the graphs. Specifically, SignGT develops a new signed self-attention mechanism (SignSA) that produces signed attention values according to the semantic relevance of node pairs. Hence, the diverse frequency information between different node pairs could be carefully preserved. Besides, SignGT proposes a structure-aware feed-forward network (SFFN) that introduces the neighborhood bias to preserve the local topology information. In this way, SignGT could learn informative node representations from both long-range dependencies and local topology information. Extensive empirical results on both node-level and graph-level tasks indicate the superiority of SignGT against state-of-the-art graph Transformers as well as advanced GNNs.