DESCRIPTIVE ANALYSIS AND COMPARISON OF REASONER USING ONTI MEASURES

Data analysis in research is an important thing to do after the research data is obtained. In designing a web application called Onti Measures, the files that have been executed have not been analyzed in more depth. Therefore, it is necessary to analyze the OWL (Web Ontology Language) files as test data for the Onti Measures application. This research aims to present a descriptive analysis of test data using three reasoners and compare their performance. The comparison of the three reasoners is seen based on running time, the performance of each reasoner, and the resulting inconsistency values. Those three reasoners are Hermit, JFact, and Pellet. In the Onti Measures application there are 10 inconsistency measures, namely drastic inconsistency measure, MI-inconsistency measure, MIc-inconsistency measure, Df-inconsistency measure, problematic inconsistency measure, incompatibility ratio inconsistency measure, MC-inconsistency measure, the nc-inconsistency measure, the mv-inconsistency measure, dan IDmcsinconsistency measure. The method used in this research is quantitative with a descriptive approach to analysis. The OWL fie input as test data is virus and disease ontology. The results of the descriptive analysis from this research include that 57.33% of the test data have an inconsistency value of 0 (consistent). Based on the performance of each reasoner in terms of processing ontologies, the three reasoners have almost the same capabilities. If it is seen from the resulting inconsistency values, the reasoner Pellet is better than the others. Meanwhile, based on the running time comparison, JFact is better than the other reasoners. The size of the ontology files does not affect the length of the running time.