W3C home > Mailing lists > Public > www-webont-wg@w3.org > July 2003

Re: daml:item - still confused

From: Jim Hendler <hendler@cs.umd.edu>
Date: Mon, 7 Jul 2003 11:55:25 -0400
Message-Id: <p05200f5fbb2f46e8ed51@[10.0.1.2]>
To: Jeff Heflin <heflin@cse.lehigh.edu>
Cc: webont <www-webont-wg@w3.org>

At 11:36 AM -0400 7/7/03, Jeff Heflin wrote:
>I think we need to cautious about adding daml:item. I can see people
>seeing it, and thinking that it provides an alternative syntax for
>building new closed lists, along the lines of:
>
><owl:Class rdf:ID="WineColor">
>     <rdfs:subClassOf rdf:resource="#WineDescriptor" />
>     <owl:oneOf>
>       <rdf:List>
>          <owl:item rdf:resource="#White" />
>          <owl:item rdf:resource="#Rose" />
>          <owl:item rdf:resource="#Red" />
>       </rdf:List>
>     </owl:oneOf>
></owl:Class>
>
>Note that many may think that this says that WineColor must be White,
>Red or Rose, but instead it simply says that that WineColor must be from
>a list that includes White, Red, Rose, and possibly others. This is very
>subtle and unintuitive, meaning it could lead to lots of
>interoperability problems down the road.
>
>However, if we decided we could live with this, wouldn't such a feature
>be better in RDF? After all, that's where the rest of the closed list
>stuff is (i.e., rdf:List, rdf:first, rdf:rest).
>
>Jeff


I would much prefer it in RDF, but they turned us down on our request 
to leave it in, and thus we would need to do it as an owl feature if 
we want it.
  I agree with the above we should be cautious about adding - but note 
that if you want to make a list that is restricted to a particular 
type in OWL as it stands now, you have to create a owl:oneOf and then 
put restrictions on the rdf first and rest properties -- which is a 
heck of a lot less intuitive than adding owl:item.
  Given Bijan offers an important use case (OWL-S) that needs this, I 
think at the very least  we should add a discuss and example to Guide 
and/or Ref a of how to make a list which is restricted to a 
particular type of item -- I don't think we have that in the docs 
right now -- Jeremy opposed adding this because we had another way to 
do it - but if we don't point out that other way, then I think Bijan 
could be justified in objecting if we don't point it out.

   -JH


-- 
Professor James Hendler				  hendler@cs.umd.edu
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)
http://www.cs.umd.edu/users/hendler      *** NOTE CHANGED CELL NUMBER ***
Received on Monday, 7 July 2003 11:55:30 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:58:01 GMT