W3C home > Mailing lists > Public > public-owl-wg@w3.org > May 2012

Re: status of xsd:duration in OWL (and RIF and SPARQL) - ACTION-164: RDF WG

From: Bijan Parsia <bparsia@cs.man.ac.uk>
Date: Tue, 8 May 2012 09:38:56 +0100
Cc: "Evain, Jean-Pierre" <evain@ebu.ch>, 'Michael Schneider' <schneid@fzi.de>, Ian Horrocks <ian.horrocks@cs.ox.ac.uk>, "public-owl-wg@w3.org" <public-owl-wg@w3.org>, "Public-Rif-Wg (E-mail)" <public-rif-wg@w3.org>, "Peter F. Patel-Schneider" <pfpschneider@gmail.com>, Sandro Hawke <sandro@w3.org>
Message-Id: <F320FB42-280C-466A-B77C-192C97C67ABA@cs.man.ac.uk>
To: Ivan Herman <ivan@w3.org>
On 8 May 2012, at 07:49, Ivan Herman wrote:
[snip]
> Jean-Pierre,
> 
> without taking side on whether duration should or should not be part of OWL 2 for reasoning (I do not have enough technical baggage to comment on that): if you use duration today, it is valid RDF (even if RDF2004 still refers to duration as a SHOULD NOT be used, but that might change now with XSD1.1). Not being part of OWL 2 means that you do not get OWL 2 datatype reasoning on those literal values; I presume an OWL 2 reasoner would treat those literal values pretty much as strings.

Nope, it would (if conforming) reject them and the surrounding ontology. Actual reasoners differ in how they handle unknown datatypes in "aggressive repair" mode.

Personally, I'd rather bypass XSD 1.1 duration and get a robust quantities datatype (so we can correctly handle unit conversion). But THAT, I agree, would violate both the letter and the spirit of the current charter :)

> So the question is what you expect from your tooling v.a.v. that particular data. 


A more robust solution would be to use a named datatype (perhaps an alias for integer) with some out of band understanding of its interpretation. You'd have to be careful about distinguishing uses of that datatype with the integers (I don't believe we have nominal type distinctions so there's no way to create an identity distinct copy of any of the datatypes, i.e., declaring integers to be disjoint with renamed integers == a contradiction; no reason this feature couldn't be introduced, fwiw).

Cheers,
Bijan.
Received on Tuesday, 8 May 2012 08:39:31 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 16:42:03 UTC