Matrix rank-based ontology matching: an extension of string equality matching

Abstract

In this paper, string-based algorithm and the analogy of the total number of deviations that exist between two matrices are utilised for ontology matching. The type of matching and the distance between them are derived from the rank of the matrix. This process is useful for an agent program to extract the knowledge from the application domain and to evaluate the result of matching to decide whether to include or exclude the ontology. As a case study to demonstrate the developed algorithm, the Fast Breeder Test Reactor (FBTR) is taken for analysis by representing its components like the steam generator system and the control rod drive mechanism in formats like Resource Description Framework (RDF) and Web Ontology Language (OWL).