Re: Error in OWL2-QL conformance test? (WebOnt-I5.26-010)

On 27/10/2010 10:10, Barry Bishop wrote:
> Hello,
>
> I do not believe that:
>
>      http://owl.semanticweb.org/page/TestCase:WebOnt-I5.26-010
>
> is a valid OWL2-QL profile test. Below is the exchange between myself
> and Michael Schneider on the subject.

Hi,

I agree with your analysis, and I find the behaviour of the OWL API 
plausible (though wrong): The OWL API follows the transformation in the 
specification, and it obtains the correct set of axioms at the end of 
the parsing (i.e. the empty set). However, the specification states that 
"At the end of this process, the graph G MUST be empty" which the OWL 
API apparently does not check (or ignores deliberately). The RDF triples 
in the result conclusion are not used in any of the RDF-to-FS 
transformation rules that are specified for OWL 2 and hence they remain 
at the end of parsing.

As I read the spec, this means that the RDF-to-FS mapping is undefined 
for this input, and hence the ontology cannot be in OWL DL. Moreover, 
all profile documents are also specified in terms of FS, so the ontology 
can not be in any profile either (since it does not have a defined FS 
representation at all).

I have changed the classification of the test accordingly and submitted 
a bug report [1] for the OWL API.

Cheers,

Markus

[1] 
https://sourceforge.net/tracker/?func=detail&aid=3098141&group_id=90989&atid=595534

>
> I hope this helps somehow. Regards,
> barry
>
> -- Barry Bishop OWLIM Product Manager Ontotext AD Tel: +43 650 2000 237
> email: barry.bishop@ontotext.com www.ontotext.com -------- Original
> Message -------- Subject: RE: Another conformance question Date: Wed, 27
> Oct 2010 10:54:22 +0200 From: Michael Schneider <schneid@fzi.de> To:
> Barry Bishop <barry.bishop@ontotext.com> CC: Spas <pachob@gmail.com> Hi
> Barry! Barry Bishop wrote:
>> >We are creating a rule implementation for OWL2-QL (for no particular
>> >reason), but have found something odd in the conformance tests:
>> >
>> >http://owl.semanticweb.org/page/TestCase:WebOnt-I5.26-010
>> >
>> >Which is odd, because OWL 2 QL specifically states that it does not
>> >support minCardinality. Is this an error in the test, i.e. that it
>> >should not apply to QL?
> Yes! While I am not so fluent with the syntactic restrictions of OWL 2
> QL, I had a look in the spec at
>
>    <http://www.w3.org/TR/2009/REC-owl2-profiles-20091027/#OWL_2_QL>
>
> and found that minCardinality restrictions are (a) explicitly excluded
> from QL in the Feature Overview (Sec. 3.1), and (b) do not appear in the
> BNF for class expressions in Sec. 3.2.3.
>
> Worse, having a closer look at the conclusion ontology, I can see that
> it is not even valid OWL 2 DL! What you see there is simply a "dangling"
> cardinality restriction, not part of any axiom.
>
> When you press klick the "Convert syntax: functional" link, then you get
> something very different, without the restriction expression anymore.
> Looks like OWL API fails to detect the error and gets entirely confused.
> Further, using the OWL 2 validator at
>
>    http://owl.cs.manchester.ac.uk/validator/
>
> claims that this is an OWL 2 QL ontology. It is not!
>
> You should report this as an error to the comment mailing list at
>
>      public-owl-comments@w3.org
>
> I suggest to attach our discussion to the mail.
>
> Cheers,
> Michael
>
> --
> Dipl.-Inform. Michael Schneider
> Research Scientist, 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
> =======================================================================
>
>
>


-- 
Markus Krötzsch
Oxford  University  Computing  Laboratory
Room 306, Parks Road, Oxford, OX1 3QD, UK
+44 (0)1865 283529    http://korrekt.org/

Received on Friday, 29 October 2010 11:48:40 UTC