W3C home > Mailing lists > Public > www-rdf-logic@w3.org > March 2003

Re: OWL Lite's restrictions on the use of the OWL vocabulary

From: Peter F. Patel-Schneider <pfps@research.bell-labs.com>
Date: Mon, 31 Mar 2003 09:10:03 -0500 (EST)
Message-Id: <20030331.091003.50030380.pfps@research.bell-labs.com>
To: yzqu@seu.edu.cn
Cc: www-rdf-logic@w3.org

From: "Yuzhong Qu" <yzqu@seu.edu.cn>
Subject: Re: OWL Lite's restrictions on the use of the OWL vocabulary
Date: Mon, 31 Mar 2003 14:22:50 +0800

> > > 8.3 OWL Lite [ In OWL Reference] says:
> > > 
> > > 1. Class axioms with an owl:equivalentClass statement. In these axioms both
> > > the domain and range should be either a class identifier or a property
> > > restriction.
> > > 
> > > 2. In value restrictions, only owl:allValuesFrom and owl:someValuesFrom may
> > > be used.(not explicitly state whether or not property-restriction type can
> > > be used )
> > > 
> > > 3. Only class descriptions of the class-identifier and property-restriction
> > > type are allowed at the righthand-side of domain and range statements for
> > > object properties.
> > > 
> > > 
> > > However. 2.3.1. OWL Lite Axioms [In OWL Abstract Syntax and Semantics]
> > > says:
> > > 
> > > A. axiom ::= 'EquivalentClasses(' classID { classID } ')'
> > > 
> > >     It means that only a class identifier can be the domain and range of an
> > > owl:equivalentClass statement.
> 
> > owl:equivalentClass triples can arise from other parts of the abstract
> > syntax, so your claim does not follow from this.
> 
> It seems that the another part of the abstract syntax is as follows:
> 
> axiom ::= 'Class(' classID modality { annotation } { super } ')'
> modality ::= 'complete' | 'partial'
> super ::= classID | restriction
> 
> Yes, it allows the range be either a class identifier or a property
> restriction.
> 
> How about the domain of an owl:equivalentClass statement (OWL Lite)?  
>
> Just classID ?  or 
> 
> classID plus restriction (as in  1/ above)?

According to S&AS, just classID.

> BTW,  Is the modality in OWL DL [2.3.2.1. OWL DL Class Axioms]
> 
> same as the modality in OWL DL [2.3.1.1. OWL Lite Class Axioms]?

Yes, is there any reason to think otherwise?

> > > B. restriction ::= ...| 'restriction(' individualvaluedPropertyID 
> > >                         { 'allValuesFrom(' classID ')'}
> > >                         { 'someValuesFrom(' classID ')' } ... ')'
> > >     
> > >     It means that only a class identifier can be used for owl:allValuesFrom
> > > and owl:someValuesFrom constructs.
> > 
> > Yes, but this is not what 2/ above says.
> 
> Thanks for your clarification.
> 
> > > C. axiom ::= 'ObjectProperty(' individualvaluedPropertyID { annotation } {
> > > 'super(' individualvaluedPropertyID ')'}
> > >               { 'domain(' classID ')'} { 'range(' classID ')' }  ... ')'
> > > 
> > >     It means that only a class identifier can be allowed at the
> > > righthand-side of domain and range statements for object properties.
> > 
> > Agreed.
> 
> Thanks again!
> 
> > > Which one is the correct statement about the OWL Lite's restrictions on the
> > > use of the OWL vocabulary?
> > 
> > Well that is a good question.  As the editor of the S&AS document, I would
> > say that this document should be definitive.
> > 
> > > Yuzhong Qu
> > 
> > peter
> >  
> 
> Yuzhong Qu
> 

peter
Received on Monday, 31 March 2003 09:10:24 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:52:43 GMT