SGVAE: Sequential Graph Variational Autoencoder

TL;DR

SGVAE introduces a sequential variational autoencoder for graphs that enhances scalability and expressivity by modeling graph construction and deconstruction as a sequential process, learning a meaningful latent space.

Contribution

The paper proposes a novel sequential graph variational autoencoder that directly models graph generation as a sequential process, improving on existing models in scalability and expressivity.

Findings

Shows promising results on cycle datasets

Highlights the need for relaxing node permutation distributions

Demonstrates the potential of sequential graph modeling

Abstract

Generative models of graphs are well-known, but many existing models are limited in scalability and expressivity. We present a novel sequential graphical variational autoencoder operating directly on graphical representations of data. In our model, the encoding and decoding of a graph as is framed as a sequential deconstruction and construction process, respectively, enabling the the learning of a latent space. Experiments on a cycle dataset show promise, but highlight the need for a relaxation of the distribution over node permutations.