W3C home > Mailing lists > Public > public-rdf-in-xhtml-tf@w3.org > September 2008

Re: lcsh.info RDFa SKOS and content negotiation - use of RDF-style # IDs in RDFa?

From: Sergey Chernyshev <rdfa.info@antispam.sergeychernyshev.com>
Date: Tue, 30 Sep 2008 18:12:49 -0400
Message-ID: <9984a7a70809301512s72db141sdca7ee94623c97a8@mail.gmail.com>
To: "Dan Brickley" <danbri@danbri.org>
Cc: RDFa <public-rdf-in-xhtml-tf@w3.org>
Dan,

I was having the same issue with RDFa and RDF/XML that were created for the
same personal profile (e.g. my profile at
http://www.sergeychernyshev.com/sergey) - the issue is not only with hash
links (#someid), but with relative (../someid) and absolute (/someid) URIs
which depend on the URL and obviously different for XHTML+RDFa document.

Although "inferencing" approaches with blank nodes and functional properties
are probably working, I chose to go with direct approach and resolve URIs
that go to XHTML+RDFa document making full URIs (e.g.
http://www.mysite.com/simeid or http://www.mysite.com/#someid in case of
hash) while keeping relative ones in RDF/XML version (being primary
storage/editable version for the application). My motivation was that first
wave on Semantic Web clients (like Search Monkey) are probably not going to
be very good at inferencing for many reasons (including performance).

I wonder if there's anything wrong with resolving the URIs to their full
form (except for maybe taking more space)?

           Sergey


On Tue, Sep 30, 2008 at 6:50 AM, Dan Brickley <danbri@danbri.org> wrote:

>
>
> Looks like http-range-14 still haunts us. I'm looking for some best
> practice advice for using "#blahblah" URI IDs in RDFa content...
>
> Backstory -
>
> Edd Summers has put up a great site at http://lcsh.info/
>
> It has one page for each concept listed in the Library of Congress Subject
> Headings, eg. http://lcsh.info/sh85112589 is a URI for a document about
> the LCSH heading "Religion and the Humanities".
>
> There is RDFa in the HTML page, as well as a complete NTriples zip-file
> download of the dataset. The RDFa uses a #concept URI to identify the
> abstract concept that the document is describing with SKOS.
>
> <div class="concept" about="http://lcsh.info/sh85112589#concept"
> typeof="skos:Concept">
>
> The document also has an RDF/XML representation available by content
> negotiation. While the triples differ a little (Ed - is this a bug?), they
> all talk about the SKOS concepts via identifiers that take the main page's
> URI and append '#concept'.
>
> Is it TAG-acceptable to use a URI in this manner, when
> http://lcsh.info/sh85112589 is the URI for a document available in RDFa
> XHTML rather than RDF/XML? What does "http://lcsh.info/sh85112589#concept"
> name? How do we 'follow our nose' here? Should the media type definition for
> RDFa allow the assignation of URIs to things other than chunks of markup?
>
> Does anyone have a suggestion for how if at all this should be done
> differently? My sense is that the HTML spec as-is doesn't license us to
> assign URIs in this way. And since #foo is to be interpreted w.r.t. to the
> (set of?) media types we get back as HTTP representations, there seems to be
> a tension here between using conneg and using rdfa and avoiding having to do
> http 303 redirects...
>
> cheers,
>
> Dan
>
>
>
> curl -s http://lcsh.info/sh85112589 | rapper -i rdfa - file:
>
> rapper: Parsing file <stdin> with parser rdfa and base URI file:
> rapper: Serializing with serializer ntriples and base URI file:
>
> <file:> <http://www.w3.org/1999/xhtml/vocab#alternate> <
> http://lcsh.info/sh85112589.rdf> .
> <file:> <http://www.w3.org/1999/xhtml/vocab#alternate> <
> http://lcsh.info/sh85112589.n3> .
> <file:> <http://www.w3.org/1999/xhtml/vocab#alternate> <
> http://lcsh.info/sh85112589.json> .
> <http://lcsh.info/sh85112589#concept> <
> http://www.w3.org/1999/02/22-rdf-syntax-ns#type> <
> http://www.w3.org/2004/02/skos/core#Concept> .
> <http://lcsh.info/sh85112589#concept> <
> http://www.w3.org/2004/02/skos/core#prefLabel> "Religion and the
> humanities" .
> <http://lcsh.info/sh85112589#concept> <
> http://www.w3.org/2004/02/skos/core#altLabel> "Humanities and religion" .
> <http://lcsh.info/sh85112589#concept> <http://purl.org/dc/terms/created>
> "1986-02-11" .
> <http://lcsh.info/sh85112589#concept> <http://purl.org/dc/terms/modified>
> "1986-02-11T00:00:00" .
> rapper: Parsing returned 8 triples
>
>
> curl -H "Accept: application/rdf+xml" -s http://lcsh.info/sh85112589 |
> rapper -i rdfxml - file:
>
> rapper: Parsing file <stdin> with parser rdfxml and base URI file:
> rapper: Serializing with serializer ntriples and base URI file:
>
> <http://lcsh.info/sh85112589#concept> <
> http://www.w3.org/2004/02/skos/core#altLabel> "Humanities and religion"@en
> .
> <http://lcsh.info/sh85112589#concept> <
> http://www.w3.org/1999/02/22-rdf-syntax-ns#type> <
> http://www.w3.org/2004/02/skos/core#Concept> .
> <http://lcsh.info/sh85112589#concept> <http://purl.org/dc/terms/created>
> "1986-02-11"^^<http://www.w3.org/2001/XMLSchema#date> .
> <http://lcsh.info/sh85112589#concept> <
> http://www.w3.org/2004/02/skos/core#prefLabel> "Religion and the
> humanities"@en .
> <http://lcsh.info/sh85112589#concept> <
> http://www.w3.org/2004/02/skos/core#broader> <
> http://lcsh.info/sh85062913#concept> .
> <http://lcsh.info/sh85112589#concept> <http://purl.org/dc/terms/modified>
> "1986-02-11T00:00:00"^^<http://www.w3.org/2001/XMLSchema#dateTime> .
> <http://lcsh.info/sh85112589#concept> <
> http://www.w3.org/2004/02/skos/core#inScheme> <http://lcsh.info/> .
> rapper: Parsing returned 7 triples
>
>
>
>


-- 
Sergey Chernyshev
http://www.sergeychernyshev.com/
Received on Tuesday, 30 September 2008 22:13:35 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 30 September 2008 22:13:39 GMT