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

Vince Jankovics, Michaël Garcia Ortiz, Eduardo Alonso

AAAI 2021 pp. 15803-15804

doi:10.1609/AAAI.V35I18.17898 /aaai/2021/jankovics2021aaai-hetsage/

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.

PDF AAAI Semantic Scholar

Cite

Text

Jankovics et al. "HetSAGE: Heterogenous Graph Neural Network for Relational Learning (Student Abstract)." AAAI Conference on Artificial Intelligence, 2021. doi:10.1609/AAAI.V35I18.17898

Markdown

[Jankovics et al. "HetSAGE: Heterogenous Graph Neural Network for Relational Learning (Student Abstract)." AAAI Conference on Artificial Intelligence, 2021.](https://mlanthology.org/aaai/2021/jankovics2021aaai-hetsage/) doi:10.1609/AAAI.V35I18.17898

BibTeX

@inproceedings{jankovics2021aaai-hetsage,
  title     = {{HetSAGE: Heterogenous Graph Neural Network for Relational Learning (Student Abstract)}},
  author    = {Jankovics, Vince and Ortiz, Michaël Garcia and Alonso, Eduardo},
  booktitle = {AAAI Conference on Artificial Intelligence},
  year      = {2021},
  pages     = {15803-15804},
  doi       = {10.1609/AAAI.V35I18.17898},
  url       = {https://mlanthology.org/aaai/2021/jankovics2021aaai-hetsage/}
}