ReGAE: Graph autoencoder based on recursive neural networks

TL;DR

ReGAE introduces a recursive neural network-based graph autoencoder capable of transforming large graphs with thousands of vertices into fixed-dimensional embeddings, enabling efficient graph operations in vector space.

Contribution

It presents a novel recursive neural network architecture for graph autoencoding that scales to large graphs, overcoming previous size limitations.

Findings

Handles graphs with thousands of vertices

Embeddings are of fixed dimension regardless of graph size

Simulation confirms effectiveness on large graphs

Abstract

Invertible transformation of large graphs into fixed dimensional vectors (embeddings) remains a challenge. Its overcoming would reduce any operation on graphs to an operation in a vector space. However, most existing methods are limited to graphs with tens of vertices. In this paper we address the above challenge with recursive neural networks - the encoder and the decoder. The encoder network transforms embeddings of subgraphs into embeddings of larger subgraphs, and eventually into the embedding of the input graph. The decoder does the opposite. The dimension of the embeddings is constant regardless of the size of the (sub)graphs. Simulation experiments presented in this paper confirm that our proposed graph autoencoder, ReGAE, can handle graphs with even thousands of vertices.