Embedding Method for Knowledge Graph with Rich Ontology

Abstract

Knowledge Graph embedding (KGEmbedding) is an expected technology to retrieve a new knowledge with completion of knowledge graph, combining of knowledge graphs and so on. However most of KG Embedding models don't consider ontology part of the knowledge graph. Especially the relationship between properties. We propose a model for knowledge graph with well-described ontology in this article. And we show some results of experiments with basic dataset and our own practical dataset.