GMNN: Graph Markov Neural Networks

TL;DR

GMNN introduces a novel neural network model that combines graph neural networks with conditional random fields, effectively capturing label dependencies for improved semi-supervised classification in relational data.

Contribution

The paper proposes GMNN, integrating GNNs with CRFs using a variational EM algorithm for joint modeling of label dependencies and object representations.

Findings

Achieves state-of-the-art results on classification tasks

Effectively models label dependencies with CRF-GNN integration

Improves semi-supervised learning performance

Abstract

This paper studies semi-supervised object classification in relational data, which is a fundamental problem in relational data modeling. The problem has been extensively studied in the literature of both statistical relational learning (e.g. relational Markov networks) and graph neural networks (e.g. graph convolutional networks). Statistical relational learning methods can effectively model the dependency of object labels through conditional random fields for collective classification, whereas graph neural networks learn effective object representations for classification through end-to-end training. In this paper, we propose the Graph Markov Neural Network (GMNN) that combines the advantages of both worlds. A GMNN models the joint distribution of object labels with a conditional random field, which can be effectively trained with the variational EM algorithm. In the E-step, one graph neural network learns effective object representations for approximating the posterior distributions of object labels. In the M-step, another graph neural network is used to model the local label dependency. Experiments on object classification, link classification, and unsupervised node representation learning show that GMNN achieves state-of-the-art results.