Benchmarking Graph Neural Networks
\nameVijay Prakash Dwivedi1 \AND\nameChaitanya K. Joshi2 \AND\nameAnh Tuan Luu1 \AND\nameThomas Laurent3 \AND\nameYoshua Bengio4 \AND\nameXavier Bresson5 \addr1Nanyang Technological University, Singapore, 2University of Cambridge, UK, 3Loyola Marymount University, USA, 4Mila, University of Montréal, Canada, 5National University of Singapore
Abstract
In the last few years, graph neural networks (GNNs) have become the standard toolkit for analyzing and learning from data on graphs. This emerging field has witnessed an extensive growth of promising techniques that have been applied with success to computer science, mathematics, biology, physics and chemistry. But for any successful field to become mainstream and reliable, benchmarks must be developed to quantify progress. This led us in March 2020 to release a benchmark framework that i) comprises of a diverse collection of mathematical and real-world graphs, ii) enables fair model comparison with the same parameter budget to identify key architectures, iii) has an open-source, easy-to-use and reproducible code infrastructure, and iv) is flexible for researchers to experiment with new theoretical ideas. As of December 2022, the GitHub repository111The framework is hosted at https://github.com/graphdeeplearning/benchmarking-gnns. has reached 2,000 stars and 380 forks, which demonstrates the utility of the proposed open-source framework through the wide usage by the GNN community. In this paper, we present an updated version of our benchmark with a concise presentation of the afore
中文速览
图神经网络（GNN）领域发展迅猛，但长期缺乏统一的基准测试框架，导致不同模型之间的比较往往不公平、不可靠。作者在2020年发布了一套开源GNN基准框架，收录了12个涵盖真实世界与数学模型的中等规模图数据集，并规定所有模型在相同参数预算下进行比较，同时提供模块化、易复现的代码基础设施，本文是对该框架的更新介绍，新增了具有真实化学溶解度测量值的分子数据集AQSOL。以该框架为实验平台，作者提出并验证了用拉普拉斯特征向量作为图位置编码（positional encoding）的思路，有效提升了消息传递类GNN在多个数据集上的表现，并由此引发了后续大量关于更强位置编码的研究。该框架已获得GitHub上2000余个星标和470余次学术引用，证明了统一基准对推动图机器学习领域可靠进步的重要价值。
原文 arXiv:2003.00982;中英对照 + 大白话阅读 https://aha.fim.ai/paper/2003.00982v5