W3C home > Mailing lists > Public > public-rdfa-wg@w3.org > September 2010

Re: Ranging Properties (was: Re: Musing on URI references vs. URI-s as string)

From: Ivan Herman <ivan@w3.org>
Date: Wed, 22 Sep 2010 12:39:03 +0200
Cc: nathan@webr3.org, W3C RDFa WG <public-rdfa-wg@w3.org>
Message-Id: <970BC864-A443-4715-8830-2BCD79640BD2@w3.org>
To: Knud Hinnerk Möller <knud.moeller@deri.org>

On Sep 22, 2010, at 12:22 , Knud Hinnerk Möller wrote:

> Hi,
> On 21 Sep 2010, at 23:26, Nathan wrote:
>> ...
>> Ranging properties:
>> Forgive my newness on this one, but has any work/discussion previously gone in to setting the type of an object based on the range of the property, for instance the range of foaf:name is foaf:Document.
> you mean foaf:homepage, right?
>> Thus it should be pretty easy to determine that if somebody writes:
>> <span property="foaf:homepage">http://www.w3.org</span>
>> Then the triple produced should be
>> :x foaf:homepage <http://www.w3.org>
>> as this is the only valid triple possible.
> As far as I know, this kind of processing - interpreting the rdfs:range of a property, or distinguishing that it was declared as an owl:ObjectProperty vs. an owl:DatatypeProperty - is happening at a higher level than just the syntactic parsing of an RDF serialisation. You would need some kind of RDFS or OWL reasoning in place (or a simple rule for cases like this). So, I'm not sure how people would react if we put something like this into RDFa Core. 

Indeed. RDFa is merely a serialization of RDF; on that level the question of the validity of that triple is not defined.

Furthermore: strictly speaking

:x foaf:homepage "http://www.w3.org" .

may almost 'valid' in RDFS sense, too! A possible rdfs:range definition on foaf:homepage is only a license to kill^H^H^H^Hinfer: a reasoner would infer that 'http://www.w3.org' is of type whatever-the-range-is. That might create other inconsistencies, of course, that might lead to invalidity, but it is further down the line, so to say.  

> On the other hand, it could be interesting to include this into an RDFa parser, to make it more robust and lenient. It would be like Web browsers accepting invalid HTML and guessing what was probably meant by it. On the one hand this might help to improve the growth of RDFa, on the other hand it might lead to chaos. 

Right and this might involve some sort of a reasoning on the client. And this is indeed interesting, but certainly not something for the RDFa spec:-)



> Cheers,
> Knud
>> Generally the aforementioned ranging approach is something which I'm unsure why we don't use more on the web of data (for both removing ambiguity wrt the type of objects, and for verifying graphs produced when parsing.
>> Still, glad to see this niggling away with somebody else :)
>> Best,
>> Nathan
> -------------------------------------------------
> Knud Möller, PhD
> +353 - 91 - 495086
> Smile Group: http://smile.deri.ie
> Digital Enterprise Research Institute
>  National University of Ireland, Galway
> Institiúid Taighde na Fiontraíochta Digití
>  Ollscoil na hÉireann, Gaillimh

Ivan Herman, W3C Semantic Web Activity Lead
Home: http://www.w3.org/People/Ivan/
mobile: +31-641044153
PGP Key: http://www.ivan-herman.net/pgpkey.html
FOAF: http://www.ivan-herman.net/foaf.rdf

Received on Wednesday, 22 September 2010 10:36:24 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:19:49 UTC