- From: Libby Miller <Libby.Miller@bristol.ac.uk>
- Date: Tue, 13 Jan 2004 14:54:28 +0000 (GMT)
- To: semantic-photolist@unitboy.com
- Cc: post@kanzaki.com, jim@jibbering.com
hi Jim, On Tue, 13 Jan 2004, Jim Hendler wrote: > Libby- > I'm forced t disagree w/a lot of the below about tools and > ontologies -- first of all, I understand your point, but if we're > supposed to have as one of our goals showing off Sem Web technology, > then it would be foolish not to use RDFS and OWL both of which are > Proposed Recs right now. If we do use them, we should use them right > -- it is pretty easy to adapt ones RDF tools to do the right thing, > and it's the right thing to do. Regardless of the exact form of the schema, I would argue only that, while we should use owl and rdfs, we should not assume that all processing tools can use them, and so we should not rely on tools being able to process them where this is unnecessary, hence my quoting Dan Connolly in my previous message. > More importantly, using equivalences, or even better just linking > via URIs is what the Sem Web is all about -- pulling eveything > together seems like a bad idea to me -- and most importantly, its not > scalable in the long run -- suppose other groups want to add other > concepts to the ontologies we are developing for this project (which > seems to me a good thing, not a bad one), they would need to add > their stuff via pointers and/or RDFS/OWL statements, or else you'd > need to let them get in and edit the main file (or otherwise assert > into it) which seems very non-SWy to me I think we're on the same side here :) In the example files I showed, I've used a number of different vocabularies, and the generated schema just links to them, giving some hints as to where it is acceptable to use each of the properties and classes. > That said, we should put some time and effort into cleaning up the > vocabularies we build this from. For example, if your stuff has a > "region" concept, and technical.owl has a "circularRegion" we should > make sure a subclass relation is put in > Looking forward to discussing this stuff on irc > -JH > p.s. to put it another way, let's not let expediency get in the way > of good design! I's say interoperability rather than 'expediency'. It's a mistake to ignore existing work *where relevant*, hence the use of geo, creative commons, dublin core, foaf and Jim Ley's SVG vocabulary in the examples I gave; however, where a vocabulary does not meet the needs of the project then it makes sense to consider designing a new one. cheers Libby ================================== This is the TEMPORARY discussion list for the W3 Semantic-Photo History Project. For questions, contact greg@fotonotes.net. Subscribe Instructions To: semantic-photolist-request@unitboy.com Body: subscribe Unsubscribe Instructions To: semantic-photolist-request@unitboy.com Body: unsubscribe Help To: semantic-photolist-request@unitboy.com Body: help
Received on Tuesday, 13 January 2004 10:01:51 UTC