I've made that mistake before as well - As a result I've largely stopped using domain and range on properties except in the case where the use of the property is very narrowly defined. I currently use restrictions for more granular control.
Tim > -----Original Message----- > From: Crapo, Andrew (GE Global Research) [mailto:[email protected]] > Sent: Tuesday, February 28, 2012 9:39 AM > To: [email protected] > Subject: RE: rdf:type inferred from rdfs:subClassOf > > Ah, my mistake. There is in the model a property "p" with range A, and > an > incorrect statement "Inst1 p B" (on which I wanted to give a warning), > from > which is concluded that "B rdf:type A". So the inference came from a > known > error in the model. > > Thanks for making me think more clearly, > Andy > > > -----Original Message----- > > From: Ian Dickinson [mailto:[email protected]] > > Sent: Monday, February 27, 2012 3:08 PM > > To: [email protected] > > Subject: Re: rdf:type inferred from rdfs:subClassOf > > > > Hi Andrew, > > On 27/02/12 19:38, Crapo, Andrew (GE Global Research) wrote: > > > Given the triple "A rdfs:subClassOf B", I'm a little puzzled that > just > > > about any Jena Reasoner Model Spec, including OWL_MEM_RDFS and > > > RDFS_MEM_RDFS will infer the triple "A rdf:type B". > > > > It shouldn't, and indeed it didn't do so when I tried just now. Can > you > supply > > a complete minimal example please? > > > > Ian
