Meta-Path Learning for Multi-Relational Graph Neural Networks

Francesco Ferrini, Antonio Longa, Andrea Passerini, Manfred Jaeger
Proceedings of the Second Learning on Graphs Conference, PMLR 231:2:1-2:17, 2024.

Abstract

Existing multi-relational graph neural networks use one of two strategies for identifying informative relations: either they reduce this problem to low-level weight learning, or they rely on handcrafted chains of relational dependencies, called meta-paths. However, the former approach faces challenges in the presence of many relations (e.g., knowledge graphs), while the latter requires substantial domain expertise to identify relevant meta-paths. In this work we propose a novel approach to learn meta-paths and meta-path GNNs that are highly accurate based on a small number of informative meta-paths. Key element of our approach is a scoring function for measuring the potential informativeness of a relation in the incremental construction of the meta-path. Our experimental evaluation shows that the approach manages to correctly identify relevant meta-paths even with a large number of relations, and substantially outperforms existing multi-relational GNNs on synthetic and real-world experiments.

Cite this Paper

BibTeX
@InProceedings{pmlr-v231-ferrini24a, title = {Meta-Path Learning for Multi-Relational Graph Neural Networks}, author = {Ferrini, Francesco and Longa, Antonio and Passerini, Andrea and Jaeger, Manfred}, booktitle = {Proceedings of the Second Learning on Graphs Conference}, pages = {2:1--2:17}, year = {2024}, editor = {Villar, Soledad and Chamberlain, Benjamin}, volume = {231}, series = {Proceedings of Machine Learning Research}, month = {27--30 Nov}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v231/ferrini24a/ferrini24a.pdf}, url = {https://proceedings.mlr.press/v231/ferrini24a.html}, abstract = {Existing multi-relational graph neural networks use one of two strategies for identifying informative relations: either they reduce this problem to low-level weight learning, or they rely on handcrafted chains of relational dependencies, called meta-paths. However, the former approach faces challenges in the presence of many relations (e.g., knowledge graphs), while the latter requires substantial domain expertise to identify relevant meta-paths. In this work we propose a novel approach to learn meta-paths and meta-path GNNs that are highly accurate based on a small number of informative meta-paths. Key element of our approach is a scoring function for measuring the potential informativeness of a relation in the incremental construction of the meta-path. Our experimental evaluation shows that the approach manages to correctly identify relevant meta-paths even with a large number of relations, and substantially outperforms existing multi-relational GNNs on synthetic and real-world experiments.} }
Endnote
%0 Conference Paper %T Meta-Path Learning for Multi-Relational Graph Neural Networks %A Francesco Ferrini %A Antonio Longa %A Andrea Passerini %A Manfred Jaeger %B Proceedings of the Second Learning on Graphs Conference %C Proceedings of Machine Learning Research %D 2024 %E Soledad Villar %E Benjamin Chamberlain %F pmlr-v231-ferrini24a %I PMLR %P 2:1--2:17 %U https://proceedings.mlr.press/v231/ferrini24a.html %V 231 %X Existing multi-relational graph neural networks use one of two strategies for identifying informative relations: either they reduce this problem to low-level weight learning, or they rely on handcrafted chains of relational dependencies, called meta-paths. However, the former approach faces challenges in the presence of many relations (e.g., knowledge graphs), while the latter requires substantial domain expertise to identify relevant meta-paths. In this work we propose a novel approach to learn meta-paths and meta-path GNNs that are highly accurate based on a small number of informative meta-paths. Key element of our approach is a scoring function for measuring the potential informativeness of a relation in the incremental construction of the meta-path. Our experimental evaluation shows that the approach manages to correctly identify relevant meta-paths even with a large number of relations, and substantially outperforms existing multi-relational GNNs on synthetic and real-world experiments.
APA
Ferrini, F., Longa, A., Passerini, A. & Jaeger, M.. (2024). Meta-Path Learning for Multi-Relational Graph Neural Networks. Proceedings of the Second Learning on Graphs Conference, in Proceedings of Machine Learning Research 231:2:1-2:17 Available from https://proceedings.mlr.press/v231/ferrini24a.html.

Related Material