- From: wangxiao <wangxiao@musc.edu>
- Date: Fri, 30 Sep 2005 16:19:45 -0400
- To: <public-semweb-lifesci@w3.org>
- Wafik > Great maybe you are right -- if 2 is wrong this means that > the over-arching layer must change when the underlying > ontologies change -- is this what is said?? No, no, no. If for example, in ontology A, it says, assuming all X, Y, Zs are owl:Class a:X owl:disjoingWith a:Y. And in Ontology B, it says b:Z rdfs:subClassOf b:Y. If your over-arching link says: a:Y owl:equivalentClass b:Y. Then the new model is still consistent. Because the inference says a:X owl:disjoint b:Z. But if you add another "arc" that says, a:X owl:equivalentClass b:Z, then there is an inconsistency. So, something must be wrong and needs to be changed. Otherwise, the description has no models to satisfy it. When a new statement is added, the model that satisfy the description most of "change". Is this the change you are talking about or "change" must be made due to "conflict"? > >> For example, if I have one ontology has protein structures and > >> another ontology has protein activity. After the > over-arching layer > >> connects both ontologies will one be able to ask the > simple question: > >> "which structure belongs to which activity"?. > > > > Of course, if you "connect" them in a meaningful way to you. > > Correct, the point is what is meaningful for one might not be > for another ... as Eric said the connection needs to be task > independant. Now if one wants to update one ontology what > will happen to the connection that is linking it to the other > ontology ... Of course, these are the engineering issues that we try to figure it out. I.e., how to create ontology in such that is evolvable and robust. > Yes, agreed the ontology must be consistent -- but how?? > Anyway, this needs a commitee or a grant to be solved and not > an email blog through recommendations, by w3c, by education, by good engineer principles, and by the willingness to adapt.
Received on Friday, 30 September 2005 20:19:58 UTC