- From: Mark van Assem <mark@cs.vu.nl>
- Date: Tue, 28 Nov 2006 15:18:53 +0100
- To: Bernard Vatant <bernard.vatant@mondeca.com>
- CC: "Williams, Stuart (HP Labs, Bristol)" <skw@hp.com>, Alistair Miles <a.j.miles@rl.ac.uk>, public-esw-thes@w3.org
Hi Bernard,
Thanks for the nice summary! Missed the middle part of the discussion
so took me a while before I understood that IFP =
inverseFunctionalProperty ;)
Another option (a variant on the last one you mention) would be to
drop subjectIndicator's IFP and instead have a reasoning rule in SKOS
that asserts skos:exactMatch between skos:subjects with the same
subjectIndicator. That protects from wrong usage of the
subjectIndicator and consequent unwanted owl:sameAs statements. Golden
rule is that if it is possible to use something wrongly it WILL happen ;)
In this solution the skos:subject values can come from existing
vocabularies, which still sounds more attractive to me than blank
nodes, because a blank node does not give any information beside the
subjectIndicator.
Another negative point for blank nodes is that in those solutions,
someone wanting to fill in a skos:subject value needs to find an
appropriate subjectIndicator first... each time s/he wants to annotate
something. A lot more work than selecting from a fixed vocabulary.
My 2 cents,
Mark.
--
Mark F.J. van Assem - Vrije Universiteit Amsterdam
markREMOVE@cs.vu.nl - http://www.cs.vu.nl/~mark
Received on Tuesday, 28 November 2006 14:19:23 UTC