- From: Deborah L. McGuinness <dlm@ksl.stanford.edu>
- Date: Sun, 02 Dec 2007 11:56:36 -0800
- To: Ian Horrocks <ian.horrocks@comlab.ox.ac.uk>
- CC: Jim Hendler <hendler@cs.rpi.edu>, OWL Working Group WG <public-owl-wg@w3.org>
Ian Horrocks wrote: > See the emails on this issue, in > particular http://lists.w3.org/Archives/Public/public-owl-wg/2007Nov/0263.html > and http://lists.w3.org/Archives/Public/public-owl-wg/2007Nov/0258.html, > for an explanation of the significant and not fully understood > technical difficulties. > > IMHO, if it goes anywhere this should go in some UFD giving advice on > "repairing" ontologies (see also ISSUE-56). this is an interesting point. there may be some number of clarification issues such as this. they may be perceived as some best practices kinds of issues. if we look at the existing model for the documents for owl 1.0, this type of advice was not really included (although some flavor of this was in the guide). this kind of information was more in the best practice working group documents. i think there is room for a range of documentation that would be useful. Are you considering suggesting some documents that are in the best practice flavor? and would those come out of this working group (or out of a swbp follow on group)? i am also wondering if we want to consider a wiki-style faq option. the reason i suggest that is if we wait for a follow on swbp group to form and documents to come out, it will be a while. thanks, deborah > > Ian > > > On 2 Dec 2007, at 18:27, Jim Hendler wrote: > >> My request in raising the issue was that if we postponed it, we >> needed a clear statement as to why the problem exists (and several >> people have suggested adding the note that many reasoners change >> rdfs:class to owl:class despite the semantic issue) -- saying >> " introduces significant and not fully understood technical >> difficulties" does not satisfy what I asked this issue to be >> addressed for. Again, in studies of ontologies out there, a great >> many OWL Full ontologies become OWL DL with this change, I would >> think that it would therefore be incumbent on this group to either >> encourage those users to use owl:class (by explaining why they >> should) or make it clear they can still work with OWL DL tools if >> they don't, but there are certain techniical risks (which I must >> admit I still don't understand) >> So if we postpone this issue, which I'm okay with, it requires an >> explanation - and that is where I think we (the OWL community) has >> failed to date. In fact, I've been suggesting to my students that >> they use rdfs:class in many applications since OWL DL tools fix it, >> and RDF-only tools don't reject it... this is clearly wrong >> semantically (Which I explain to the best of my limited ability) but >> right from a practical point of view... >> -JH >> >> >> >> >> On Dec 1, 2007, at 3:07 PM, Ian Horrocks wrote: >> >>> >>> I'm tending to postpone this issue on the grounds that it introduces >>> significant and not fully understood technical difficulties. >>> >> >> "If we knew what we were doing, it wouldn't be called research, would >> it?." - Albert Einstein >> >> Prof James Hendler http://www.cs.rpi.edu/~hendler >> <http://www.cs.rpi.edu/%7Ehendler> >> Tetherless World Constellation Chair >> Computer Science Dept >> Rensselaer Polytechnic Institute, Troy NY 12180 >> >> >> >> >
Received on Sunday, 2 December 2007 19:56:54 UTC