Lists Home |
Date Index |
Dave Pawson wrote:
> Isn't is necessary for the same grouping to agree terminology?
OWL provides the semantics of "subclassOf", "equivalentProperty",
"sameIndividualAs", etc. Ontology creators must define the terms and
their relationships. The idea I espouse is to reach agreement on the
logical model and encourage diversity of physical expressions.
> I guess this is where the human aspects come in?
Those who define the ontology specifies the terms and their semantics.
Note that OWL is a "Web" technology. Thus, an Ontology may evolve in a
distributed fashion (just like the Web itself!)
For example, suppose that when I build my application the Camera
Ontology only defines these terms:
Camera, aperture, (lens) size
I then construct my application to understand physical expressions (XML
instance documents) which uses these terms. Now, suppose that time
passes and the Camera Ontology evolves to include these additional
SLR, f-stop, focal-length
Further, these relationships are specified in the Ontology:
"SLR is a type of Camera"
"f-stop is synonymous with aperture"
"focal-length is synonymous with (lens) size"
Now, without any modifications to my application, I can process this
How are this be? After all, it is using terms (SLR, f-stop, and
focal-length) that my application was not constructed to understand.
Well, when my application encounters a term that it does not understand
it "consults" the Camera ontology:
"What do you know about SLR?"
The Ontology returns:
"SLR is a subclassOf of Camera"
My application understands:
- "subclassOf" since it's part of the OWL vocabulary
- "Camera" since my application was constructed to understand this
So, my application now understands that this physical expression is
talking about Cameras. Further, it is talking about a particular type
Thus, without a-priori agreement my application is able to process a
trading partner's document! /Roger