Re: [protege-user] Arguments for and against OWL
Dave McComb via protege-user <[email protected]>
| Newsgroups | gmane.comp.misc.ontology.protege.owl |
|---|---|
| Message-ID | <[email protected]> |
These arguments have been roiling for decades. Our opinion at Semantic Arts is that OWL is useful. A few elaborations, and qualifications: * Very few people in an organization need to understand owl, and even fewer need to be able to author it well, but everyone will benefit. * The benefit is having an enterprise ontology that is as elegant (simple) as possible and as unambiguous as possible * Once you have an elegant/ unambiguous model you can re-express it in SHACL, UML, JSON-LD, ERDs whatever, but this ability to re-express it in a different modeling paradigm tends to cause people to think that it wasn’t necessary, and that they might have created as elegant a model in one of these other languages and disciplines. * Perhaps it a bit like saying if someone created a work of art in oil paint, and then took a picture of it and pasted it in powerpoint, and then started editing it, you could say you didn’t really need the oil paint or the artistic ability, because you could “just do everything in powerpoint.” Just because you can create some stuff in powerpoint and edit some stuff does not mean it is the only medium or skill you need. * Think of owl like oil paints, most people are never going to get very good with them. Think of ontology modeling as artistry, even fewer are going to master this skill. * What we have found is that the discipline of creating full formal models, from a limited number of primitive classes and properties, forces modelers to get far beyond the surface meaning and into the deep meaning of concepts, where the real opportunity to find the similarities and differences lie. Every reduction in complexity, that does not suffer from lack of precision is a big gain. Many orders of magnitude reduction in complexity are routine, but every percentage reduces the effort for everything else that follows. * We run inference in design, only occasionally in production. But this discipline finds many logical errors, and forces one to be clear and complete in ones thinking and communicating. See gist for what we think is a good example, and starting point https://www.semanticarts.com/gist/ On Jan 21, 2025, at 5:31 PM, Ann Clark via protege-user <[email protected]> wrote: I'm seeing some noise online (LinkedIn) arguing against using OWL. The main argument seems to be that OWL is not human readable, by people who do not appear to have learned OWL. Has anyone on this list got a good list of pros and cons for using OWL for applied ontology? Or, any articles arguing for OWL? Thank you in advance. Ann Get Outlook for Android<https://aka.ms/AAb9ysg> _______________________________________________ protege-user mailing list [email protected] https://mailman.stanford.edu/mailman/listinfo/protege-user _______________________________________________ protege-user mailing list [email protected] https://mailman.stanford.edu/mailman/listinfo/protege-user