RE: Use of owl:versionInfo to record ontology version IRI in RDF

Hi!

Quick Summary: I don't think that anything needs to be changed.
 
Here is the actual text from the OWL 1 Reference:

<http://www.w3.org/TR/owl-ref/#versionInfo-def>

[[
     An owl:versionInfo statement 
-1-> generally has as its object a string 
     giving information about this version, for example RCS/CVS keywords. 
     This statement does not contribute 
     to the logical meaning of the ontology other than 
-2-> that given by the RDF(S) model theory.
]]

>From -1- it sounds to me that this is only a very informative statement.

Now to -2-: "no other meaning than that by RDFS semantics" means to me that
the range of owl:versionInfo is rdfs:Resource, i.e. can be anything. The
reason is that this property isn't part of the RDF(S) vocabulary, so it is
just a plain property for RDFS. And this means that, if an RDFS reasoner
sees a triple "s owl:versionInfo o", then it cannot conclude other than "o
rdf:type rdfs:Resource".

To strengthen this: In OWL 1 Full (er, the OWL 1 RDF-Compatible Semantics
:)), the only thing that is said about owl:versionInfo was that it is an
owl:AnnotationProperty (see Section 5.2 of
<http://www.w3.org/TR/owl-semantics/rdfs.html>). And for an
owl:AnnotationProperty p the range was determined to be owl:Thing union
rdfs:Literal (Table "Characteristics of OWL classes, datatypes, and
properties" in Section 5.2). This was, in OWL 1 Full, just equivalent to
owl:Thing, and this was again equivalent to rdfs:Resource, i.e. everything.
 
Btw, in the OWL 2 RDF-Based Semantics, I have explicitly defined the range
of owl:versionInfo to be unrestricted, just to be clear. :)

So, I don't see an issue here. Maybe in common use owl:VersionInfo has often
been used with a string as its object (e.g. with the "RCS/CVS keywords"
mentioned in the OWL Ref). But I cannot see anything in the (normative
part!) of the OWL 1 spec that speaks against using IRIs as objects as well.

Michael

>-----Original Message-----
>From: public-owl-wg-request@w3.org [mailto:public-owl-wg-request@w3.org]
>On Behalf Of Uli Sattler
>Sent: Thursday, April 02, 2009 1:14 PM
>To: W3C OWL Working Group
>Subject: Fwd: Use of owl:versionInfo to record ontology version IRI in
>RDF
>
>Dear OWL,
>
>Matthew Horridge found an 'unwanted feature' in the current spec when
>implementing it in the OWL API [1], see message below. This was
>discussed with Boris, who suggested:
>
>"I see that http://www.w3.org/TR/owl-ref/#versionInfo-def says that the
>value
>should be a string. I guess this wouldn't be that difficult to change:
>we could
>have something like owl:versionURI in OWL 2. "
>
>
>I guess we should do this sooner rather than later?
>
>Cheers, Uli
>
>[1] http://owlapi.sourceforge.net/
>
>Begin forwarded message:
>
>
>	From: Matthew Horridge <matthew.horridge@cs.man.ac.uk>
>	Date: 1 April 2009 13:48:37 BST
>	To: Uli Sattler <sattler@cs.man.ac.uk>
>	Subject: Use of owl:versionInfo to record ontology version IRI in
>RDF
>
>	Hi,
>
>	I realised that the OWL API wasn't translating an ontology's
>version IRI into RDF, and I was about to fix this.  However, I saw that
>the current mapping uses owl:versionInfo from OWL 1 to encode this.  I
>just wondered whether or not another property was considered.  The
>reason is that the OWL 1 spec states that the value of the versionInfo
>property is a string.  Because of this, it might not be possible to
>parse the versionInfo property of an existing ontology into an IRI.
>There might also be several versionInfo annotations on an ontology, and
>in this case it's not clear which one to choose.
>
>	As an example of an ontology that would be difficult to parse
>correctly, consider the pizza ontology at
>
>	http://www.co-ode.org/ontologies/pizza/2007/02/12/pizza.owl
>
>	Rightly or wrongly, this ontology has three versionInfo
>annotations and all of them are general comments about what went into
>successive versions of the ontology.  I've seen other ontologies like
>this as well.
>
>	Would it be possible (if it's not too late, wouldn't cause too
>much trouble etc.) to coin a new piece of vocab to store the version URI
>of an ontology?  Something like  ontologyVersion?  Also, would it be
>possible to specify what to do when there are multiple version IRIs?
>
>	Cheers,
>
>	Matthew
>
>	------------------------------------------------------------------
>------------------------------------------------------------------------
>-----
>	7.4.1 owl:versionInfo
>	An owl:versionInfo statement generally has as its object a string
>giving information about this version, for example RCS/CVS keywords.
>This statement does not contribute to the logical meaning of the
>ontology other than that given by the RDF(S) model theory.
>
>	Although this property is typically used to make statements about
>ontologies, it may be applied to any OWL construct. For example, one
>could attach a owl:versionInfo statement to an OWL class.
>
>	NOTE: owl:versionInfo is an instance of owl:AnnotationProperty.
>
>

--
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
=======================================================================