Using agents to parallelize a medical reasoning system based on ontologies and description logics as an application case

• Published in: Expert systems with applications Vol. 39; no. 18; pp. 13085 - 13092
• Main Authors: Rodríguez-González, Alejandro, Torres-Niño, Javier, Hernández-Chan, Gandhi, Jiménez-Domingo, Enrique, Alvarez-Rodríguez, Jose Maria
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 15.12.2012
• Subjects: Architecture; Design engineering; Distributed; Expert system; Expert systems; Logic; Medical; Medical diagnosis; Ontologies; Parallel; Platforms; Reasoning; Semantic web; Tasks; Semantic web; Reasoning; Expert system; Distributed; Ontologies; Parallel; Medical diagnosis
• ISSN: 0957-4174; 1873-6793
• DOI: 10.1016/j.eswa.2012.05.093

► Multi-Agent architectures and Distribute Reasoning improves inference times. ► Reasoning over Description Logics needs reasoners with high-scalability capabilities. ► The proposed architecture can deal with scalability issues. ► Parallelization and multithreading can improve multi-agent architecture results. Reasoning over ontologies is a hard task where the efficiency in terms of time usually depends on the ontology expressivity and the reasoner used. Depending on the domain and the techniques used to design the ontology, this fact could be a decisive factor to decide if the reasoning platform is useful or not. In this paper we describe an application case of a multi-agent architecture to parallelize a medical diagnosis system based on ontologies and description logics. The aim of this paper is to show the improvement of the time efficiency of the system in which is based this work using the proposed architecture. Results show that the use of the new architecture (which is based in the use of a master-slave architecture and the use of multi-threading) shows that global efficiency can be highly improved.