W3C home > Mailing lists > Public > public-powderwg@w3.org > November 2008

Discussion of foaf:depiction/wdrs:logo (was Updates to POWDER-S voc and namespace doc)

From: Phil Archer <parcher@fosi.org>
Date: Thu, 27 Nov 2008 14:59:46 +0000
Message-ID: <492EB5E2.4010103@fosi.org>
To: Public POWDER <public-powderwg@w3.org>
CC: Dan Brickley <danbri@danbri.org>

I'm forwarding this edited version of an e-mail sent to the member list 
by Andrea in October, following a comment on the blog from Dan Bri [1]

I _think_ it answers his point?

[1] http://www.w3.org/blog/powder/2008/11/25/meeting_summary_13_october_2008

-------- Original Message --------
Subject: Updates to POWDER-S voc and namespace doc
Resent-Date: Mon, 13 Oct 2008 10:49:11 +0000
Resent-From: member-powderwg@w3.org
Date: Mon, 13 Oct 2008 12:48:24 +0200
From: Andrea Perego <andrea.perego@uninsubria.it>
To: Member POWDER <member-powderwg@w3.org>

As agreed, I've performed the following changes (to the POWDER-S 
vocabulary http://www.w3.org/2007/05/powder-s):
- wdrs:Document is defined as a subclass of owl:Ontology
- wdrs:issuedby is defined as an owl:AnnotationProperty, without any
range or subproperty specification

Moreover, since any property used in owl:Ontology should be either an
annotation property or an ontology property, the following has been
defined as annotation properties:
- validFrom
- validUntil
- data_error

All of them have either an rdfs:domain or rdfs:range constraint.
However, since they don't use any OWL feature for domain/range
definition, this shouldn't be a problem (Stasinos?).

In addition to the ones above, since dcterms:issued is supposed to be
used for the issue date, also this one has been included in the voc, and
defined as an annotation property.

Note that this concerns a more general issue, which, if I'm not
mistaken, might affect the DR doc [...]

The problem is the following. So far, we have discussed about annotation
properties only for the attribution part of a DR. However, they apply
not only to OWL ontology headers, but also to OWL class definitions
(which is how IRI set definitions and the descriptorset parts of DRs are
mapped into). More precisely, in OWL DL, all properties, except the ones
used to define a class (i.e., rdfs:subClassOf, owl:unionOf,
owl:intersectionOf, etc.), MUST be annotation properties. Quoting an
example from the OWL Reference [1]:

<owl:AnnotationProperty rdf:about="&dc;creator"/>

<owl:Class rdf:about="#MusicalWork">
     <rdfs:label>Musical work</rdfs:label>
     <dc:creator>N.N.</dc:creator>
</owl:Class>

As you can see, here dc:creator is used in a class definition only after
it has been defined as an owl:AnnotationProperty.

As far as POWDER is concerned, this means that, if we want to map
wdr:displaytext and wdr:displayicon to dcterms:description and
foaf:depiction, respectively (see the DR doc [2]), the latter must be
defined in the POWDER-S voc as annotation properties. As far as
dcterms:description is concerned, this is not a problem. However, this
does not work with foaf:depiction, which in the FOAF voc is defined as
an owl:ObjectProperty. So, we have two possible options:
- replace foaf:depiction with another suitable property defined in an
existing voc
- define a new property

[1]http://www.w3.org/TR/owl-ref/#Annotations
[2]http://www.w3.org/TR/2008/WD-powder-dr-20080815/#operational
Received on Thursday, 27 November 2008 15:00:20 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:42:13 GMT