Improving the Expressive Power of Graph Neural Network with Tinhofer Algorithm

TL;DR

This paper introduces WLT-GNN, a novel graph neural network variation inspired by Tinhofer's work, which enhances expressive power beyond the Weisfeiler-Lehman test limitations, supported by theoretical insights and empirical results.

Contribution

The paper proposes WLT-GNN, a new message passing scheme that theoretically surpasses WL test limitations, improving GNN expressive power.

Findings

WLT-GNN demonstrates comparable performance to existing models.

WLT-GNN shows improved expressive power on benchmark datasets.

Theoretical analysis confirms breaking WL test limitations.

Abstract

In recent years, Graph Neural Network (GNN) has bloomly progressed for its power in processing graph-based data. Most GNNs follow a message passing scheme, and their expressive power is mathematically limited by the discriminative ability of the Weisfeiler-Lehman (WL) test. Following Tinhofer's research on compact graphs, we propose a variation of the message passing scheme, called the Weisfeiler-Lehman-Tinhofer GNN (WLT-GNN), that theoretically breaks through the limitation of the WL test. In addition, we conduct comparative experiments and ablation studies on several well-known datasets. The results show that the proposed methods have comparable performances and better expressive power on these datasets.