- From: Kendall Grant Clark <kendall@clarkparsia.com>
- Date: Tue, 15 Apr 2008 09:24:30 -0400
- To: John Goodwin <john.goodwin@ordnancesurvey.co.uk>
- Cc: "Fink, Clayton R." <clayton.fink@jhuapl.edu>, public-owl-dev <public-owl-dev@w3.org>
Excerpts from John Goodwin's message of Tue Apr 15 06:17:14 -0400 2008: > Thanks for your reply Clay, however you really don't want to trust > ontologies those Ordnance Survey people write...they have no idea what > they are talking about [1] > > > > I think there might be two issues here - one is doing RCC8 type > reasoning within OWL, i.e. reasoning about "part of", "touches" etc. and > the other is how do we (in a standard way) reference and use spatial > datatypes in OWL/RDF. I know the latter may be a way off yet, but it > would be nice to start thinking about it. It would certainly be nice to > see OWL instance stores/triplestores in the future being able to handle > spatial datatypes in an analogous way that spatial databases can. John, et. al., We have some plans to extend our DL-Lite reasoner, Owlgres, with a temporal logic, DLR, which I think we could use to provide spatio-temporal reasoning of the sort you refer to, John. This would give -- hand-waving a bit here -- the RCC8 operators for spatio-temporal units. I'm being purposefully vague here because I haven't read the DLR paper carefully yet and because I'm not entirely sure how this would work. But, in principle, Owlgres would retain its nice computational tractability even with these extensions. But, again, hand-wave, hand-wave...Someone who works with me who really understands this stuff will have to say more about it. :> -- Cheers, Kendall
Received on Tuesday, 15 April 2008 13:25:09 UTC