Over the last decade, several ontology reasoners have been proposed to overcome the computational complexity of inference tasks on expressive ontology languages such as OWL 2 DL. Nevertheless, it is well-accepted that there is no outstanding reasoner that can outperform in all input ontologies. Thus, deciding the most suitable reasoner for an ontology based application is still a time and effort consuming task. In this paper, we suggest to develop a new system to provide user support when looking for guidance over ontology reasoners. At first, we will be looking at automatically predict a single reasoner empirical performances, in particular its robustness and efficiency, over any given ontology. Later, we aim at ranking a set of candidate reasoners in a most preferred order by taking into account information regarding their predicted performances. We conducted extensive experiments covering over 2500 well selected real-world ontologies and six state-of-the-art of the most performing reasoners. Our primary prediction and ranking results are encouraging and witnessing the potential benefits of our approach.
CITATION STYLE
Alaya, N., Ben Yahia, S., & Lamolle, M. (2016). Ranking with ties of OWL ontology reasoners based on learned performances. In Communications in Computer and Information Science (Vol. 631, pp. 234–259). Springer Verlag. https://doi.org/10.1007/978-3-319-52758-1_14
Mendeley helps you to discover research relevant for your work.