Re: [OWLAPI-developer] using RDF lists in object property assertions

[Jim B. <ba...@gm...>]

Thanks! It’s helpful to know that it wouldn’t necessarily be rejected. It does bother me that it is very clearly unsupported. I’ll have to see how critical depending on RDF lists is to the members of the working group.

Thanks,
Jim


> On Mar 3, 2022, at 10:06 AM, Ignazio Palmisano <ipa...@gm...> wrote:
> 
> Hi,
> Declaration of rdf:first as an object property would create profile violations, but whether that's acceptable or not depends on the application - OWLAPI won't mind too much.
> 
> I'm not sure if altering the parser to do this is doable - it's certainly complicated, that part of the code is quite convoluted. We'd accept a contribution to implement this depending on what its effects are; it's not something supported by the specs, so the issue really is that we should be sure it doesn't affect users with compliant ontologies and does not create a (new) maintenance problem.
> 
> Cheers,
> I.
> 
> On Thu, 3 Mar 2022, 14:33 Jim Balhoff, <ba...@gm... <mailto:ba...@gm...>> wrote:
> Hello,
> 
> I’ve recently been participating in the FHIR RDF working group, which is revising the RDF representation for data in the FHIR data model (https://www.fhir.org <https://www.fhir.org/>). One of the desires of the group, from an RDF perspective, is to represent ordered data using standard RDF lists (i.e., using rdf:first and rdf:rest, etc.). But they also would like FHIR datasets to be usable as OWL. I know that according to the OWL spec, using lists in this way is forbidden. But I did some experiments on the results of declaring rdf:first and rdf:rest as object properties, and adding a few helper axioms to support accessing items in lists. If such an ontology is written in OWL functional syntax, OWL API seems to work as expected, including various reasoners. It even writes the ontology to RDF “correctly”. But parsing such an ontology from RDF doesn’t work. All the abox axioms using lists disappear. I made a summary document: https://docs.google.com/document/d/1kNkP0EHfUiEj9GwO4NtzHXQD4kVTViphV0ngGfHBnBI/edit# <https://docs.google.com/document/d/1kNkP0EHfUiEj9GwO4NtzHXQD4kVTViphV0ngGfHBnBI/edit#>
> 
> Is there any way that the RDF-to-OWL parser could allow these axioms to be parsed? If it could be done, would OWL API accept a pull request making this possible? I’ve thought of a few other alternatives, such as providing a SPARQL Update which could be run prior to using a dataset as OWL. But this creates a hurdle to working with OWL which we’d like to avoid.
> 
> Best regards,
> Jim
> 
> ____________________________________________
> James P. Balhoff, Ph.D.
> Renaissance Computing Institute (RENCI)
> University of North Carolina at Chapel Hill
> 
> 
> 
> _______________________________________________
> Owlapi-developer mailing list
> Owl...@li... <mailto:Owl...@li...>
> https://lists.sourceforge.net/lists/listinfo/owlapi-developer <https://lists.sourceforge.net/lists/listinfo/owlapi-developer>
> _______________________________________________
> Owlapi-developer mailing list
> Owl...@li...
> https://lists.sourceforge.net/lists/listinfo/owlapi-developer