Motif-based Convolutional Neural Network on Graphs

TL;DR

This paper presents Motif-CNN, a graph neural network that uses motif-based spatial convolution and attention to effectively capture high-order structural information, achieving significant improvements in node classification tasks.

Contribution

It introduces a novel motif-based convolution operation and an attention mechanism for heterogeneous graphs, advancing graph neural network capabilities.

Findings

Achieved 6-21% accuracy improvements over existing methods.

Effectively captures high-order structural information.

Demonstrated on real-world social and heterogeneous graph datasets.

Abstract

This paper introduces a generalization of Convolutional Neural Networks (CNNs) to graphs with irregular linkage structures, especially heterogeneous graphs with typed nodes and schemas. We propose a novel spatial convolution operation to model the key properties of local connectivity and translation invariance, using high-order connection patterns or motifs. We develop a novel deep architecture Motif-CNN that employs an attention model to combine the features extracted from multiple patterns, thus effectively capturing high-order structural and feature information. Our experiments on semi-supervised node classification on real-world social networks and multiple representative heterogeneous graph datasets indicate significant gains of 6-21% over existing graph CNNs and other state-of-the-art techniques.