W3C home > Mailing lists > Public > public-owl-wg@w3.org > June 2009

Proposal to make certain ontology properties of OWL 2 Full into annotation properties

From: Michael Schneider <schneid@fzi.de>
Date: Mon, 22 Jun 2009 14:21:26 +0200
Message-ID: <0EF30CAA69519C4CB91D01481AEA06A0014C5A8F@judith.fzi.de>
To: "W3C OWL Working Group" <public-owl-wg@w3.org>
Hi!

While creating a testsuite for OWL 2 RL and OWL 2 Full, I realized that there is probably a mismatch between the two languages regarding annotation properties, which would formally lead to unsoundness of the RL ruleset w.r.t. the RDF-Based Semantics. 

In Section 4.3 of the Profiles Spec, rule "prp-ap" of Table 5 results in a triple 

    ap rdf:type owl:AnnotationProperty

"for every built-in annotation property ap of OWL 2 RL".

I guess that these "built-in annotation properties" are meant to be those listed in Section 5.5 of the Structural Spec. Now, under the RDF-Based Semantics several of the built-in annotation properties listed in the Structural Spec are /not/ in class owl:AnnotationProperty, but are of type owl:OntologyProperty (an example is owl:priorVersion). This mismatch has been discussed in the past (in the discussion about the QRG), but hasn't yet lead to real technical issues, so I did not see any need for action so far. But now, I think something needs to be done. 

Note, however, that just changing the ontology properties into annotation properties would formally break backwards compatibility of OWL 2 Full, which would be a bad idea for a CR document. So I propose a more conservative change.

PROPOSED: All ontology properties of the RDF-Based Semantics that are listed as annotation properties in Sec. 5.5 of the Structural Specification are /additionally/ made into annotation properties (instances of class owl:AnnotationProperty) in the RDF-Based Semantics. 

This change would just be enough to fix the concrete technical problem, and would also be enough to bring the so far different notions of annotation properties between DL and Full in line, most probably without having any unforeseeable semantic side effects. 

I would consider this just to be a "typical bugfix" for a Call-for-Implementation phase. It's a small issue, and the problem actually occurred when I was trying to "implement" testcases that work both for Full and RL.

This change would exclusively have an effect on the RDF-Based Semantics document. (Although maybe the QRG can be reworded a bit after the change, since there would not be any need to talk about ontology properties anymore. This has to be checked by the authors of the QRG.)

I suggest to vote on this change on the next TC.

Best,
Michael

--
Dipl.-Inform. Michael Schneider
Research Scientist, Dept. Information Process Engineering (IPE)
Tel  : +49-721-9654-726
Fax  : +49-721-9654-727
Email: michael.schneider@fzi.de
WWW  : http://www.fzi.de/michael.schneider
=======================================================================
FZI Forschungszentrum Informatik an der Universität Karlsruhe
Haid-und-Neu-Str. 10-14, D-76131 Karlsruhe
Tel.: +49-721-9654-0, Fax: +49-721-9654-959
Stiftung des bürgerlichen Rechts, Az 14-0563.1, RP Karlsruhe
Vorstand: Prof. Dr.-Ing. Rüdiger Dillmann, Dipl. Wi.-Ing. Michael Flor,
Prof. Dr. Dr. h.c. Wolffried Stucky, Prof. Dr. Rudi Studer
Vorsitzender des Kuratoriums: Ministerialdirigent Günther Leßnerkraus
=======================================================================




Received on Monday, 22 June 2009 12:22:03 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 22 June 2009 12:22:04 GMT