- From: Alan Rector <rector@cs.man.ac.uk>
- Date: Thu, 06 Jan 2005 10:41:51 +0000
- To: NANNI Marco RD-BIZZ-SOP <marco.nanni@francetelecom.com>
- CC: public-swbp-wg@w3.org
- Message-ID: <41DD15EF.CEA120B9@cs.man.ac.uk>
Marco I don't think everybody needs to be able to solve every puzzle, but people do need to be aware of the pitfalls of which this is an example, but probably in the ways they are likely to come up in real life modelling. In our tutorials we try to cover - although not in these words and with meaningful examples * Closure axioms and the need for them * That "only does not imply some" - that you almost always need to an existential restriction somewhere, and that in most ontologies, existential restrictions are the default (Probably the most common error we see is to use universal restrictions where existential restrictions are required) * Trivial satisfiability - that C is satisfiable * The fact that a trivially satisfiable restriction is subsumed by any non-trivially satisfiable restriction - that D subsumes C * Also the effect of disjoint axioms on the above. Our debugging tools also look for trivially satisfiable universal restrictions because they are almost always errors, most commonly arising from the use of AND instead of OR between disjoint entities, but sometimes for less obvious reasons. They don't find all possible cases, but they find most of the ones people actually make. They'll be on the CO-ODE web site "real soon now". Regards Alan NANNI Marco RD-BIZZ-SOP wrote: > Hi, > > First of all happy new year to all. > > Here 2 definitions to illustrate my questions : > > C = (and (all R P) (all R (not P) ) > D= (and (all R P') (all R (at-most 4 S)) > > I have to admit that all the OWL newcomers and even people who have already > built some not only little academicals examples of OWL ontologies to which i > have showed such examples dont understand that : > > - C is satifiable (consistent ?) > - D subsume C. > It'is very diificult to explain such (classification) results. > > So my questions are ? > > 1) do you think that "ordinary" people MUST be capable to clearly > explain themselves these examples or is this kind of case > > only for experts of the domain (the member of this WG - not me of > course) ? The problem is that i can't find a "real life" (with concrete > classes names) to show that if, i don't understand this point i could make > some mistake in creating "real" ontologies. In other words is the example - or > the fact to not understand it - the visible part of a huge Iceberg or the > symptom of a deeper incomprehension problem for potentials mistakes in using > OWL regularly > > 2) do you think that such points are covered by the tutorial material > in http://www.w3.org/2001/sw/BestPractices/Tutorials ? > > 3) do you think it is in the scope of the WG ? > > Thanks a lot > > Best regards > > Marco NANNI -- Alan L Rector Professor of Medical Informatics Department of Computer Science University of Manchester Manchester M13 9PL, UK TEL: +44-161-275-6188/6149/7183 FAX: +44-161-275-6236/6204 Room: 2.88a, Kilburn Building email: rector@cs.man.ac.uk web: www.cs.man.ac.uk/mig www.opengalen.org www.clinical-escience.org www.co-ode.org
Received on Thursday, 6 January 2005 10:40:45 UTC