Rules versus OWL: Evaluated!

About a decade ago, not long after description logics had been chosen as the basis for the then-forthcoming W3C Recommendation for the Web Ontology Language OWL, a rather agressively voiced discussion as to whether a rule-based paradigm might have been a better choice emerged in the Semantic Web community. One of the arguments brought forth was that it were much easier to convey logical statements using rules rather then OWL (or description logic) axioms, thus making it easier for ontology engineers to model. 

To the best of my knowledge, this claim was never really evaluated. The discussion around rules versus OWL also subsided significantly in the years since: OWL simply became the de-facto prominent paradigm for ontology modeling, and furthermore research results showed that there is a tighter relationship between OWL and rules than seemed the case during the past heated discussions.

In particular, a body of work showed how a lot of rules can be converted into OWL, and this brought us back to the earlier (unverified) argument that modeling with rules were easier than modeling directly with OWL axioms. 

As a consequence, we have developed a Protégé plugin called ROWLTab that allows users to enter OWL axioms by way of rules; the plugin then automatically converts these rules into OWL 2 DL axioms if possible, and prompts the user in case such a conversion is not possible without weakening the semantics of the rule. We furthermore used ROWLTab to evaluate the claim made earlier, and indeed, our data shows that modeling with rules (using ROWLTab) is significantly quicker and less prone to errors than modeling without it, for hard modeling tasks. 

To find out more, please consult our paper  at ESWC 2017 about this. You can also check out the ROWLTab plug-in website and our evaluation page. Any feedback is most welcome.

This work was supported by the NSF through the TROn project.