Are we really making much progress? Revisiting, benchmarking, and refining heterogeneous graph neural networks

TL;DR

This paper critically re-evaluates the progress of heterogeneous graph neural networks, revealing that simple homogeneous GNNs often outperform complex HGNNs when properly configured, and introduces a standardized benchmark and a strong baseline for future research.

Contribution

It systematically reproduces recent HGNNs, uncovers underestimated performance of simple GNNs, and provides a standardized benchmark and a new strong baseline to advance HGNN research.

Findings

Simple GNNs like GCN and GAT can outperform HGNNs with proper settings.

The Heterogeneous Graph Benchmark (HGB) standardizes evaluation across datasets.

The proposed Simple-HGN baseline outperforms previous models on HGB.

Abstract

Heterogeneous graph neural networks (HGNNs) have been blossoming in recent years, but the unique data processing and evaluation setups used by each work obstruct a full understanding of their advancements. In this work, we present a systematical reproduction of 12 recent HGNNs by using their official codes, datasets, settings, and hyperparameters, revealing surprising findings about the progress of HGNNs. We find that the simple homogeneous GNNs, e.g., GCN and GAT, are largely underestimated due to improper settings. GAT with proper inputs can generally match or outperform all existing HGNNs across various scenarios. To facilitate robust and reproducible HGNN research, we construct the Heterogeneous Graph Benchmark (HGB), consisting of 11 diverse datasets with three tasks. HGB standardizes the process of heterogeneous graph data splits, feature processing, and performance evaluation. Finally, we introduce a simple but very strong baseline Simple-HGN--which significantly outperforms all previous models on HGB--to accelerate the advancement of HGNNs in the future.