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

Re: Last Call comments on RDFa Core

From: Harry Halpin <hhalpin@w3.org>
Date: Mon, 13 Dec 2010 12:36:06 -0000 (GMT)
Message-ID: <245d37b681f46e10235af891dbd7b195.squirrel@webmail-mit.w3.org>
To: "Toby Inkster" <tai@g5n.co.uk>
Cc: "Harry Halpin" <hhalpin@w3.org>, "Ivan Herman" <ivan@w3.org>, public-rdfa-wg@w3.org
> On Mon, 2010-12-13 at 11:19 +0000, Harry Halpin wrote:
>> To my knowledge, I have not seen a single halfway convincing usecase
>> where there is a reason why you would want to 'mention' a URI, i.e.
>> refer to it as a literal or xsd string.
>
> Given the following:
>
> ###
> @prefix con: <http://www.w3.org/2000/10/swap/pim/contact#> .
> @prefix owl: <http://www.w3.org/2002/07/owl#> .
>
> <http://www.w3.org/People/Berners-Lee/card#i>
> 	con:preferredURI <http://www.w3.org/People/Berners-Lee/card#i> ;
> 	owl:sameAs <http://identi.ca/user/45563> .
> ###
>
> An OWL-capable processor can make the following conclusion:
>
> ###
> @prefix con: <http://www.w3.org/2000/10/swap/pim/contact#> .
>
> <http://www.w3.org/People/Berners-Lee/card#i>
> 	con:preferredURI <http://identi.ca/user/45563> .
> ###
>
> However, given:
>
> ###
> @prefix con: <http://www.w3.org/2000/10/swap/pim/contact#> .
> @prefix owl: <http://www.w3.org/2002/07/owl#> .
>
> <http://www.w3.org/People/Berners-Lee/card#i>
> 	con:preferredURI "http://www.w3.org/People/Berners-Lee/card#i" ;
> 	owl:sameAs <http://identi.ca/user/45563> .
> ###
>
> It will not come to a mistaken conclusion about what timbl's preferred
> URI for himself is.


Yes, TimBL also gave me that example. However, that example is not
plausible, much less convicing.

It seems to be the fault of a misuse of sameas and the attendant wrong OWL
conclusions, again, of which I have done an analysis of, feel free to read
[1].  The right answer is probably not to use sameAs or OWL at all, but
use something like a SIOC term for accountOf.  Right now we are doing a
study of how most things that are inferred via using OWL over sameas are
wrong, our guess so far is more than half of the inferences that use
sameas are incorrect. Given that sameAs usage is often broken, usecases
involving sameAs are IMHO red herrings.

So, try again. Not convincing a reason for essentially doubly the number
of prefix terms in RDFa and making widespread vocabularies like OGP not
interpreted as they were intended.

[1] http://iswc2010.semanticweb.org/pdf/261.pdf

>
> --
> Toby A Inkster
> <mailto:mail@tobyinkster.co.uk>
> <http://tobyinkster.co.uk>
>
>
Received on Monday, 13 December 2010 12:36:09 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:19:50 UTC