Cart
[Home] [By Thread] [By Date] [Recent Entries]
Thomas B. Passin wrote: > > Walter, even though I agree with your remarks generally, I do not think that > Roger's work here is as far off the mark as you say. Your counter example > of a Customs application is good and right on the point, but consider this. > If the Customs application knows what a "Camera" is - according to both > ontologies - then it probably has enough info to decide if the thing is a > dutiable good or not by consulting its own expertise. If it can use the SLR > ontology to bounce from SRL up to Camera, and __if__ it can find out that > the SLR's Camera is equivalent to the Custom's Camera, then the app is all > set. There are several ways to do this: 1) The Customs ontology uses the SLR ontology "camera" class directly using its URI. e.g. http://example.org/CameraOntology#camera 2) One can explicitly state the logical equivalence between two classes (the sets have identical members) as: customs:Camera owl:sameClassAs camera:camera . 3) With the appropriate "surrounding information" an OWL inferencing engine can *infer* that the classes are identical. i.e. given a bunch of triples/statements the OWL engine spits out => customs:Camera owl:sameClassAs camera:camera . > > The key is not for each ontology to clutter up the other (as you suggestesd > might have to happen), but to establish the equivalence. OTOH, I personally > think that is the hard part! One problem among many is to how to know when > two terms are in fact "equivalent" since that will usually imply a lot of > background knowledge that may or may not be articulated or shared. > Right, or just use the term directly and there is no question that customs is talking about a camera. That's the point of shared ontologies: you don't need to redefine everything in every ontology. Jonathan
|
