HyperKG: Hyperbolic Knowledge Graph Embeddings for Knowledge Base Completion
Learning embeddings of entities and relations existing in knowledge bases allows the discovery of hidden patterns not captured in the original data. In this paper, we propose a novel embedding model, dubbed HyperKG, for knowledge base completion. Our model extends translational models by exploiting hyperbolic space in order to better reflect the topological properties of knowledge bases. We investigate the type of regularities that our model can capture and we show that HyperKG is a prominent candidate for effectively representing a subset of Datalog rules. We empirically show, using a variety of link prediction datasets, that HyperKG overcomes the performance barriers of Euclidean translational models narrowing the performance gap against the bilinear models.
READ FULL TEXT