W3C home > Mailing lists > Public > www-webont-wg@w3.org > March 2002

Re: LANG: OWL non-xml syntax

From: Jonathan Borden <jonathan@openhealth.org>
Date: Tue, 26 Mar 2002 17:22:37 -0500
Message-ID: <050501c1d514$bdad54c0$0a2e249b@nemc.org>
To: "Frank van Harmelen" <Frank.van.Harmelen@cs.vu.nl>, <www-webont-wg@w3.org>
Frank,

Since what I am suggesting has not been as clear as I would like, and since
I haven't received a peep on
http://lists.w3.org/Archives/Public/www-webont-wg/2002Mar/0292.html, let me
translate the question into DAML+OIL, so that it might be more
understandable. I would also like to know whether people agree that this is
a problem (my hope is that I've gotten something else mistaken, and someone
can explain why this isn't a problem)

I raised what might be a problematic class definition that uses "oneOf" to
define a class of individuals (and note that in this case "oneOf" is
intended :-)

class ex:foo (
        oneOf(
                individual  (
                            uniquelyIdentifyingProperty <#mbox>
(mailto:connolly@w3.org)
                            property <#noseColor> (<#green>)
                            )
                individual (
                            uniquelyIdentifyingProperty <#mbox>
(mailto:connolly@w3.org)
                            property <#member> (w3c:WOWG)
                )
        ...
        )
    )

now lets assume that each individual can be given an individualID ,e.g.

individual ex:ind1 (
                            uniquelyIdentifyingProperty
<#mbox>(mailto:connolly@w3.org)
                            property <#noseColor> (<#green>)
                            )
individual ex:ind2 (
                            uniquelyIdentifyingProperty <#mbox>
(mailto:connolly@w3.org)
                            property <#member> (w3c:WOWG)
                            )
class ex:foo (
        oneOf(  individual ex:ind1,
                      invividual ex:ind2  )
    )

[for this reason I _do_ suggest that the production describing "oneOf" allow
"individual" rather than "individualID"]

and in DAML+OIL, translating an "individual" into an rdf:Description:

<rdf:Description rdf:ID="ind1">
     <ex:mbox rdf:resource="mailto:connolly@w3.org"/>
     <ex:noseColor rdf:resource="#green"/>
</rdf:Description>

<rdf:Description rdf:ID="ind2">
     <ex:mbox rdf:resource="mailto:connolly@w3.org"/>
     <ex:member rdf:resource="http://www.w3.org/2001/sw/WebOnt"/>
</rdf:Description>

<Class rdf:ID="foo">
    <oneOf rdf:parseType="daml:collection">
            <Thing rdf:resource="#ind1"/>
            <Thing rdf:resource="#ind2"/>
    </oneOf>
</Class>

The question: what does the above mean in RDF, what is it intended to mean
in OWL?

To me, the RDF interpretation asserts that Dan Connolly has a green nose.
To me, the OWL interpretation states that there exists a class whose members
are two individuals. The _problem_ is that we can't state a class of
individuals without asserting the individuals _unless_ some other syntax is
used to 'describe' individuals -- but what would that look like?

If I haven't correctly translated into DAML+OIL, then say so, and we can
decide _how_ the proposed OWL abstract syntax _ought to be_ translated into
DAML+OIL. But without having an unambiguous syntax that we can write down
and discuss, I don't see how we can make these decisions.

Jonathan
Received on Tuesday, 26 March 2002 17:25:42 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:57:48 GMT