A unified framework on node classification using graph convolutional networks

Abstract

Graphs contain a plethora of valuable information about the underlying data which can be extracted, analyzed, and visualized using Machine Learning (ML). The challenge is that graphs are non-Euclidean structures, and cannot be directly used with ML techniques. In order to overcome this challenge, one way is to encode nodes into an equivalent Euclidean representation in the form of a low-dimensional vector, also called an embedding vector, and the encoding process is called node embedding. During the recent years, various ML techniques have been developed that learn the encoding of the nodes automatically. Some of these techniques, called Graph Convolutional Networks (GCN), use variants of the convolutional neural networks adapted for graphs. The focus of this paper is two-fold. Firstly, to develop a unified framework focusing on three major GCN techniques in order to analyze, evaluate, and compare their performance on select benchmark datasets for the task of node classification. And secondly, to implement a new attention aggregator for GraphSAGE, and compare the performance of the aggregator with the existing GCN methods as well as the other aggregators provided by GraphSAGE.