A Theory of Link Prediction via Relational Weisfeiler-Leman on Knowledge Graphs
Abstract
Graph neural networks are prominent models for representation learning over graph-structured data. While the capabilities and limitations of these models are well-understood for simple graphs, our understanding remains incomplete in the context of knowledge graphs. Our goal is to provide a systematic understanding of the landscape of graph neural networks for knowledge graphs pertaining to the prominent task of link prediction. Our analysis entails a unifying perspective on seemingly unrelated models and unlocks a series of other models. The expressive power of various models is characterized via a corresponding relational Weisfeiler-Leman algorithm. This analysis is extended to provide a precise logical characterization of the class of functions captured by a class of graph neural networks. The theoretical findings presented in this paper explain the benefits of some widely employed practical design choices, which are validated empirically.
Keywords
Cite
@article{arxiv.2302.02209,
title = {A Theory of Link Prediction via Relational Weisfeiler-Leman on Knowledge Graphs},
author = {Xingyue Huang and Miguel Romero Orth and İsmail İlkan Ceylan and Pablo Barceló},
journal= {arXiv preprint arXiv:2302.02209},
year = {2023}
}
Comments
Proceedings of the Thirty-Seventh Annual Conference on Advances in Neural Information Processing Systems (NeurIPS 2023). Code available at: https://github.com/HxyScotthuang/CMPNN