RE: Planned changes to the VIAF RDF

Ed,

The problem is that authority data makes a distinction between
established and variant headings that should be preserved somehow in
authority Linked Data. If we assume skos:prefLabel/skos:altLabel are the
best solution for this, then we need to contend with the SKOS S14
integrity condition:

<http://www.w3.org/TR/skos-reference/#S14>

SKOS tries to preserve S14 integrity for pure skos:Concepts (e.g. LCSH)
by discouraging owl:sameAs and using skos:exactMatch instead:

"owl:sameAs, owl:equivalentClass or owl:equivalentProperty would
typically be inappropriate for linking SKOS concepts in different
concept schemes, because the formal consequences that follow could be
undesirable." <http://www.w3.org/TR/skos-reference/#L4858>

Even though there is no rdfs:domain set on skos:prefLabel and they COULD
be assigned as properties of people, it's not practical to discourage
owl:sameAs for those types of entities. This means that all the
skos:prefLabels get thrown into an owl:sameAs pile and S14 could be
inadvertently violated. The foaf:focus patterns solves this. 

In other words, the skos:Concept/foaf:focus pattern is being used in
this case to control *headings*, not because the #concept has other
roles to play such as in dcterms:subject. For those cases, use the
#thing URI instead.

I admit this use of skos:Concept/foaf:focus/owl:Thing to control
headings for non-conceptual things is confusing, but as a design pattern
is it very nice. 

Jeff

> -----Original Message-----
> From: public-lld-request@w3.org [mailto:public-lld-request@w3.org] On
> Behalf Of Ed Summers
> Sent: Wednesday, April 13, 2011 8:51 AM
> To: public-lld@w3.org
> Subject: Re: Planned changes to the VIAF RDF
> 
> Hi Jeff,
> 
> First, let me just say I'm a big fan of the simplifications that you
> and Thom are proposing ... they are clearly a big improvement. But I
> am wondering about the foaf:focus pattern that you are promoting.
> 
> I know I've said this before privately in IRC to various people, but
> it's probably worth asking aloud here. Is it really necessary to use
> URIs to distinguish between the thing itself, and the concept of the
> thing? If we do, how would we like to see these resources used in RDF
> descriptions?
> 
> For example, if loc.gov were to follow your advice [1] and assert
that:
> 
>     <http://id.loc.gov/authorities/n88102106#concept> a skos:Concept ;
>         skos:prefLabel "Tillett, Barbara B." ;
>         foaf:focus <http://id.loc.gov/authorities/n88102106#thing> .
> 
>     <http://id.loc.gov/authorities/n88102106#thing> a foaf:Person .
> 
> Would you want to see an assertion that Barbara Tillett wrote
> "Bibliographic relationships : toward a conceptual structure of
> bibliographic information used in cataloging" (assuming the book URI
> was legit) done like this:
> 
>     <http://www.worldcat.org/oclc/17832769#book> dcterms:creator
> <http://id.loc.gov/authorities/n88102106#thing> .
> 
> And an assertion that the article "An Interview with Barbara B.
> Tillett" is about Barbara Tillett would be:
> 
>     <http://www.worldcat.org/oclc/440932158#article> dcterms:subject
> <http://id.loc.gov/authorities/n88102106#concept> .
> 
> I think lesser minds (at least mine) often have difficulty knowing
> when to use a URI for the Information Resource (aka Document) in their
> descriptions as opposed to a URI for the Non-Information Resource (aka
> Real World Thing). I think further distinguishing between the Concept
> of the Thing and the Real World Thing (both as Real World Things?)
> actually compounds the problem.
> 
> Personally, I would prefer to see loc.gov only mint one identifier for
> Barbara Tillett:
> 
> <http://id.loc.gov/authorities/n88102106#i> a foaf:Person ;
>         foaf:name "Tillett, Barbara B." .
> 
> And let people use that identifier in their descriptions:
> 
>     <http://www.worldcat.org/oclc/17832769#book> dcterms:creator
> <http://id.loc.gov/authorities/n88102106#i> .
> 
>     <http://www.worldcat.org/oclc/440932158#article> dcterms:subject
> <http://id.loc.gov/authorities/n88102106#i> .
> 
> I'm generally worried about this tendency of proliferating URIs, in
> the absence of guidance on how the URIs are to be used. I also think
> that the motivations for minting separate resources can often be
> appeased by refined notions of how the resource URIs are to be used in
> descriptions [2].
> 
> //Ed
> 
> [1] http://www.mail-archive.com/code4lib@listserv.nd.edu/msg10116.html
> [2] http://dfdf.inesc-id.pt/tr/web-arch
> 
> On Tue, Apr 12, 2011 at 10:21 AM, Young,Jeff (OR) <jyoung@oclc.org>
> wrote:
> > Thom Hickey posted a blog entry about our plans to streamline the
> VIAF RDF.
> >
> >
> >
> > http://outgoing.typepad.com/outgoing/2011/04/changes-to-viafs-
> rdf.html
> >
> >
> >
> > I can elaborate on the listserv if anyone wants to discuss the
> changes.
> >
> >
> >
> > Jeff
> >
> >
> >
> > ---
> >
> > Jeffrey A. Young
> > Software Architect
> > OCLC Research, Mail Code 410
> > OCLC Online Computer Library Center, Inc.
> > 6565 Kilgour Place
> > Dublin, OH 43017-3395
> > www.oclc.org
> >
> > Voice: 614-764-4342
> > Voice: 800-848-5878, ext. 4342
> > Fax: 614-718-7477
> > Email: jyoung@oclc.org
> >
> >
>

Received on Wednesday, 13 April 2011 13:51:33 UTC