Hypergraphs provide a natural representation for many real world datasets. We propose a novel framework, HNHN, for hypergraph representation learning. HNHN is a hypergraph convolution network with nonlinear activation functions applied to both hypernodes and hyperedges, combined with a normalization scheme that can flexibly adjust the importance of high-cardinality hyperedges and high-degree vertices depending on the dataset. We demonstrate improved performance of HNHN in both classification accuracy and speed on real world datasets when compared to state of the art methods.
@article{arxiv.2006.12278,
title = {HNHN: Hypergraph Networks with Hyperedge Neurons},
author = {Yihe Dong and Will Sawin and Yoshua Bengio},
journal= {arXiv preprint arXiv:2006.12278},
year = {2020}
}