W3C home > Mailing lists > Public > public-owl-dev@w3.org > January to March 2007

Re: AllDisjoint in RDF mapping

From: Jim Hendler <hendler@cs.umd.edu>
Date: Wed, 21 Feb 2007 23:45:24 -0500
Message-Id: <p06240801c202cfb90312@[]>
To: Matthew Horridge <matthew.horridge@cs.man.ac.uk>, public-owl-dev@w3.org

At 5:41 PM +0000 2/20/07, Matthew Horridge wrote:
>The functional spec allows a disjoint classes axiom to contain 2 or 
>more class descriptions.  As far as I can tell, in the current RDF 
>mapping, disjoint classes must be mapped in a pairwise fashion using 
>disjointWith statements.  In some cases, for example when following 
>the design pattern of making primitive sibling classes mutually 
>disjoint, this can cause the size of the RDF graph to become very 
>large.  Additionally, there is no way to preserve "modelling style" 
>because the current mapping results in round tripping problems after 
>serialising and parsing.
>To solve these problems, would it be possible to add an AllDisjoint 
>mapping to the spec (rather like AllDifferent for individuals)?

Let me join Matthew in requesting this - note also that 
"DisjointUnion" doesn't answer this need in all cases - often we want 
to make things disjoint that either belong to many different classes 
or that we have disjunction of classes without wanting to imply that 
it is in some way complete.
How can you be in two places at once if you're not anywhere at all?
						(Firesign Theatre, 1969)
Prof James Hendler				http://www.cs.rpi.edu/~hendler
Tetherless World Constellation Chair		301-405-2696 (work)
Computer Science Dept			301-405-6707 (Fax)
Rensselaer Polytechnic Institute, Troy NY 12180
Received on Thursday, 22 February 2007 04:53:14 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 20:58:14 UTC