Substructure Discovery in Commonsense Relations Using Graph Representation Learning

Acquiring commonsense knowledge and reasoning is an important goal in modern natural language processing research. Despite much progress, there is still a lack of understanding (especially at scale) of the nature of commonsense knowledge itself. A potential source of structured commonsense knowledge that could be used to derive insights is ConceptNet. In particular, ConceptNet contains several coarse-grained relations, including ‘HasContext’, ‘FormOf’ and ‘SymbolOf’, which can prove invaluable in understanding broad, but critically important, commonsense notions such as ‘context’. In this article, we present a methodology based on unsupervised knowledge graph representation learning and clustering to reveal and study substructures in three coarse-grained and heavily used relations in ConceptNet. Our results show that, despite having an ‘official’ definition in ConceptNet, many of these coarse-grained relations exhibit considerable substructure. In the future, therefore, such relations could be sub-divided into other relations with more refined definitions. We also supplement our core study with visualizations and qualitative analyses.