Re: OWL issue: constraining members of a container in OWL:DL

Stephane - apologies for not getting back to you sooner - I wanted to 
check w/the RDF Core folks to make sure we hadn't miscommunicated 
with them.   I think the situation is pretty clear -- as Jeremy 
Carroll put it in a note the webont mailing list [1]

In terms of the comment on RDF, I think we only really claim good
interoperability between OWL Full and RDF. Between OWL DL and RDF some
things work and some don't.

and the situation is that your current encoding is not in OWL DL. 
You go on to say

" seems that it is not possible to do this in OWL-DL. Correct me if I
am wrong. I think this is major flaw in OWL-DL if it is not supported
because this scenario can occur many times."

but I don't really understand that.  OWL DL is a *limited subset* 
(technically, a "profile')  of OWL and thus there should be no 
expectation that it can handle all the things that may occur in using 
the Owl vocabularies.   OWL itself handles the situation below just 
fine - it's just that it goes outside the limits of OWL DL in terms 
of what can be expressed in a provably decidable way.   Many things 
will do that, and  many OWL tools will support those things just 
fine.  In fact, section 1.4 of the OWL Reference [2] states:

RDF documents will generally be in OWL Full, unless they are 
specifically constructed to be in OWL DL or Lite.


  Section 8 of that document is a discussion of the relation between 
the three OWL sublanguages (Lite, DL, and Full) which explains our 
restrictions and the purpose of them

I hope this addresses your problem, and I apologize for taking so 
long to get back to you
  -Jim Hendler
p.s. If you wish to continue this discussion w/respect to how you 
might rework your examples to be in OWL DL, a good place for that is 
the www-rdf-logic@w3.org mailing list, which is intended for  general 
discussion of these sorts of issues.




[1] http://lists.w3.org/Archives/Public/www-webont-wg/2004Feb/0026.html
[2] http://www.w3.org/TR/2004/REC-owl-ref-20040210



At 17:34 -0500 2/17/04, Stephane Fellah wrote:
>Hi,
>
>I am currently trying to define some rules to convert UML to OWL. For
>property of cardinality n and for which order is important, I want to
>use rdf:Seq. I would like to constraint the member of the rdf:Seq to a
>specific type (such as xs:String or a particular Class).
>
>Example: Here how an Address is expressed in ISO 19115.
>
>
><rdf:RDF
>xmlns:iso19115="http://www.pcigeomatics.com/ontology/iso/19115.owl#"
>           xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#" >
>  <rdf:Address rdf:nodeID="urn:pcigeomatics:hull-office">
>     <iso19115:deliveryPoint>
>       <rdf:Seq>
>         <rdf:li>PCI Geomatics Inc</rdf:li>
>         <rdf:li>490, Boulevard St Joseph</rdf:li>
>         <rdf:li>Suite 400</rdf:li>
>       </rdf:Seq>
>     </iso19115:deliveryPoint>
>     <iso19115:postalCode>J8Y  3Y7</iso19115:postalCode>
>     <iso19115:city>Hull</iso19115:city>
>     <iso19115:administrativeArea>Quebec</iso19115:administrativeArea>
>     <iso19115:country>Canada</iso19115:country>
>   </rdf:Address>
></rdf:RDF>
>
>The order of the deliveryPoint address is important in this case (think
>about printing a label for the address).I would like to express in
>OWL-DL that deliveryPoint property is rdf:Seq has take only xs:String or
>rdfs:Literal for its member.
>
>It seems that it is not possible to do this in OWL-DL. Correct me if I
>am wrong. I think this is major flaw in OWL-DL if it is not supported
>because this scenario can occur many times.
>
>Note that this issue has been raised in RDFS discussion and has been
>forwarded to the ontology WG. It seems that this issue has not been
>addressed. Ref:
>http://www.w3.org/2000/03/rdf-tracking/#rdfs-constraining-containers.
>
>Any suggestions to solve this problem are welcome.
>
>Best regards
>
>Stephane Fellah
>Senior Software Engineer
>
>PCI Geomatics
>490, Boulevard St Joseph
>Hull, Quebec
>Canada J8Y 3Y7
>Tel: 1 819 770 0022 Ext. 223
>Fax 1 819 770 0098
>Visit our web site:  www.pcigeomatics.com
>

-- 
Professor James Hendler			  http://www.cs.umd.edu/users/hendler 
Director, Semantic Web and Agent Technologies	  301-405-2696
Maryland Information and Network Dynamics Lab.	  301-405-6707 (Fax)
Univ of Maryland, College Park, MD 20742	  240-277-3388 (Cell)