Improving Join Query Performance for Linked Data Using Knowledge Graph Embedding

Abstract

Recently, Semantic Web data are characterized by exponential growth in size and changing data. For the use and processing of Semantic Web data, Linked Data based on RDF (Resource Desciprion Framework) must be done first. The use of simple hash values for RDF data has the disadvantage of having a large amount of space complexity. In this paper, join queries of Linked Data are experimented using various knowledge graph techniques. The performance of these methods is compared and analyzed with existing methods. Experimental results confirmed that knowledge graph models perform better than the existing hash method and the TransD model is best.