HetSAGE: Heterogenous Graph Neural Network for Relational Learning (Student Abstract)

Authors

  • Vince Jankovics Artificial Intelligence Research Centre (CitAI), City, University of London
  • Michael Garcia Ortiz Artificial Intelligence Research Centre (CitAI), City, University of London
  • Eduardo Alonso Artificial Intelligence Research Centre (CitAI), City, University of London

Keywords:

Neuro-symbolic, Graph Neural Network, Inductive Logic Programming, Inductive Learning, Knowledge Graphs, Relational Learning, Heterogeneous Graphs

Abstract

This paper aims to bridge this gap between neuro-symbolic learning (NSL) and graph neural networks (GNN) approaches and provide a comparative study. We argue that the natural evolution of NSL leads to GNNs, while the logic programming foundations of NSL can bring powerful tools to improve the way information is represented and pre-processed for the GNN. In order to make this comparison, we propose HetSAGE, a GNN architecture that can efficiently deal with the resulting heterogeneous graphs that represent typical NSL learning problems. We show that our approach outperforms the state-of-the-art on 3 NSL tasks: CORA, MUTA188 and MovieLens.

Downloads

Published

2021-05-18

How to Cite

Jankovics, V., Garcia Ortiz, M., & Alonso, E. (2021). HetSAGE: Heterogenous Graph Neural Network for Relational Learning (Student Abstract). Proceedings of the AAAI Conference on Artificial Intelligence, 35(18), 15803-15804. Retrieved from https://ojs.aaai.org/index.php/AAAI/article/view/17898

Issue

Section

AAAI Student Abstract and Poster Program