Abstract:

Reasoners are one of the main components of the ontological systems and the work of reasoners is the most resource-intensive task in ontology processing. The study proposes the reasoner combination method to enhance the performance of ontological systems. Its essence is selecting the most suitable reasoner out of the HermiT, Pellet and FaCT++ reasoners depending on the type of ontology. The distinctive feature of the research is combining the advantages of tableau and hypertableau methodologies. The criterion, which has been developed, allows you to choose a reasoner for an ontology with optimal performance based on ontology components: TBox, ABox, RBox. The results of the studies clearly show that the application of the reasoner combination method outstrips the performance of any particular reasoner, considering that the reasoner will process different types of ontology. The testing method was conducted on a set of 8 different ontologies: BP XP OBOL, FMA Lite Fly Taxonomy, Biological Process, DLP ExtDnS, MGED, DOLCE-Plans, SWEET Numerics. The study resulted in the development of the method which will allow applying ontologies in the tasks that require high performance and processing of a large body of knowledge. Author Biographies Igor Bibichkov, Kharkiv National University of Radio Electronics Nauky ave., 14, Kharkiv, Ukraine, 61166 Postgraduate student Department of Artificial Intelligence

Keywords: