Dan Brickley | 7 Jan 2010 00:00

Re: [foaf-dev] FOAFs use of owl:AnnotationProperty

Hi Andrew,

On Wed, Jan 6, 2010 at 11:27 PM, Houghton,Andrew <houghtoa@...> wrote:
> I have a question about FOAF's use of owl:AnnotationProperty for DCMI properties.  It seems to me that if
I load the FOAF RDF, the SKOS RDF and some SKOS concepts that use DCMI properties into a triple store.  Any
DCMI properties in my SKOS concepts, specifically dc:description, dc:title and dc:date, will be
treated as owl:AnnotationProperty rather than properties because the FOAF RDF declares them as such
which is not what was intended.
>
> So my question is should FOAF be placing this constraint on other ontologies?

No, it shouldn't. In fact this is one of those features of OWL that
really annoyed me; that DC title etc was unusable in lots of DL
environments without telling the little lie that it's an annotation
propertly. Nevertheless there has been a lot of work in the OWL DC
scene and I didn't want to pointlessly diverge from compatibility, so
I grudgingly put those lies into the schema:

See the RDF/XML comments for detail -  http://xmlns.com/foaf/spec/index.rdf

  <!--  DC terms are NOT annotation properties in general, so we
consider the following
	claims scoped to this document. They may be removed in future
revisions if OWL tools become more flexible. -->
  <owl:AnnotationProperty
rdf:about="http://purl.org/dc/elements/1.1/description"/>
  <owl:AnnotationProperty rdf:about="http://purl.org/dc/elements/1.1/title"/>
  <owl:AnnotationProperty rdf:about="http://purl.org/dc/elements/1.1/date"/>

Since we make such minimalistic use of dc: there anyway, it would
(Continue reading)

Houghton,Andrew | 7 Jan 2010 00:21
Picon
Favicon

Re: [foaf-dev] FOAFs use of owl:AnnotationProperty

> From: Dan Brickley [mailto:danbri@...]
> Sent: Wednesday, January 06, 2010 06:01 PM
> To: Houghton,Andrew
> Cc: foaf-dev Friend of a
> Subject: Re: [foaf-dev] FOAFs use of owl:AnnotationProperty
> 
> > So my question is should FOAF be placing this constraint on other
> ontologies?
> 
> No, it shouldn't. In fact this is one of those features of OWL that
> really annoyed me; that DC title etc was unusable in lots of DL
> environments without telling the little lie that it's an annotation
> propertly. Nevertheless there has been a lot of work in the OWL DC
> scene and I didn't want to pointlessly diverge from compatibility, so
> I grudgingly put those lies into the schema:
> 
> Since we make such minimalistic use of dc: there anyway, it would
> perhaps be better to omit dc: completely.
>
> ... and use foaf:name instead of dc:title, or perhaps rdfs:label,
> rdfs:comment?

It might be better to omit those DCMI annotation properties and use
rdfs:label, rdfs:comment and/or appropriate SKOS annotation properties.

However, I just realized that FOAF was using the DCMI elements
namespace while my SKOS concepts were using the DCMI terms namespace.
This might not be an issue, I'm not sure, since the 15 original DCMI
elements in the DCMI terms namespace are defined as rdfs:subPropertyOf
rather than owl:sameAs.  Which in this case could be a good thing.  I
(Continue reading)


Gmane