Article Classification with Graph Neural Networks and Multigraphs
This addresses the problem of classifying research articles into taxonomies for researchers and librarians, but it is incremental as it builds on existing GNN methods with multi-graph enhancements.
The paper tackles article classification by enhancing Graph Neural Networks with multi-graph representations that encode multiple signals of article relatedness, resulting in consistent performance improvements across datasets and enabling simple pipelines to match complex architectures.
Classifying research output into context-specific label taxonomies is a challenging and relevant downstream task, given the volume of existing and newly published articles. We propose a method to enhance the performance of article classification by enriching simple Graph Neural Network (GNN) pipelines with multi-graph representations that simultaneously encode multiple signals of article relatedness, e.g. references, co-authorship, shared publication source, shared subject headings, as distinct edge types. Fully supervised transductive node classification experiments are conducted on the Open Graph Benchmark OGBN-arXiv dataset and the PubMed diabetes dataset, augmented with additional metadata from Microsoft Academic Graph and PubMed Central, respectively. The results demonstrate that multi-graphs consistently improve the performance of a variety of GNN models compared to the default graphs. When deployed with SOTA textual node embedding methods, the transformed multi-graphs enable simple and shallow 2-layer GNN pipelines to achieve results on par with more complex architectures.