Relational Graph Attention Networks

Abstract

We investigate Relational Graph Attention Networks, a class of models that extends non-relational graph attention mechanisms to incorporate relational information, opening up these methods to a wider variety of problems. A thorough evaluation of these models is performed, and comparisons are made against established benchmarks. To provide a meaningful comparison, we retrain Relational Graph Convolutional Networks, the spectral counterpart of Relational Graph Attention Networks, and evaluate them under the same conditions. We find that Relational Graph Attention Networks perform worse than anticipated, although some configurations are marginally beneficial for modelling molecular properties. We provide insights as to why this may be, and suggest both modifications to evaluation strategies, as well as directions to investigate for future work.

10 pages + 8 pages of appendices. Layer implementation available at https://github.com/Babylonpartners/rgat/

References in corpus (9)

• Semi-Supervised Classification with Graph Convolutional Networks
• Geometric deep learning: going beyond Euclidean data
• Molecular Graph Convolutions: Moving Beyond Fingerprints
• Deep Convolutional Networks on Graph-Structured Data
• Convolutional Networks on Graphs for Learning Molecular Fingerprints
• Massively Multitask Networks for Drug Discovery
• GaAN: Gated Attention Networks for Learning on Large and Spatiotemporal Graphs
• Attention Models in Graphs: A Survey
• Dual-Primal Graph Convolutional Networks

Cited by in corpus (18)

• Graph Neural Networks: Taxonomy, Advances and Trends
• Semi-Supervised Variational Reasoning for Medical Dialogue Generation
• GNN-FiLM: Graph Neural Networks with Feature-wise Linear Modulation
• How Attentive are Graph Attention Networks?
• Relational Graph Convolutional Networks: A Closer Look
• Bag of Tricks for Node Classification with Graph Neural Networks
• Entity Alignment with Reliable Path Reasoning and Relation-Aware Heterogeneous Graph Transformer
• Automatic Intersection Management in Mixed Traffic Using Reinforcement Learning and Graph Neural Networks
• r-GAT: Relational Graph Attention Network for Multi-Relational Graphs
• TeMP: Temporal Message Passing for Temporal Knowledge Graph Completion
• Graph Warp Module: an Auxiliary Module for Boosting the Power of Graph Neural Networks in Molecular Graph Analysis
• Learning to Execute Programs with Instruction Pointer Attention Graph Neural Networks
• Graph Ensemble Learning over Multiple Dependency Trees for Aspect-level Sentiment Classification
• Weisfeiler-Lehman Embedding for Molecular Graph Neural Networks
• Improving the Long-Range Performance of Gated Graph Neural Networks
• Neural Trees for Learning on Graphs
• Node Attribute Completion in Knowledge Graphs with Multi-Relational Propagation
• Edge-similarity-aware Graph Neural Networks