Hi Stuart,
> Antoine, I am uncertain about:
>
> Another point mentioned in the minutes: as good practice an ontology shouldn't re-define elements from an existing ontology, as APs do sometimes for the vocabulary they re-use.
> But an ontology can certainly refine by means of introducing sub-classes and sub-properties of an existing ontology.
> And in fact the issue with "local re-definition" might be once again a matter of available representation techniques. In the future, one could imagine ontologies with named graphs that include ontological statements on already-defined classes and properties. That would match I think this "local interpretation" that is done via APs.
>
> Antoine
>
>
> It has been my understanding that an AP "shouldn't re-define elements" either and that refinement needs for an AP follow the same sub-class/sub-property mechanisms. In the purest sense, application profiles shouldn't "define" properties at all but only local constraints on already defined properties--i.e., properties defined outside the AP itself.
Yes, I agree. I've tried to use "local re-definition" later to distinguish the AP-style local constraints from a true re-definition case. Looks like I've perfectly illustrated the kind of issues that may arise if we assume a too specific interpretation of the word "ontology", which was the starting point ;-)
Antoine
> Thus if my AP uses properties from schemas A and B created by others and I need some new properties to meet the needs defined by the AP, I can create schema C (which may include refinements of properties in schemas A & B) that can then be "used" by the AP.
>
> Stuart
> --
> Stuart A. Sutton
> CEO, Dublin Core Metadata Initiative
> Associate Professor Emeritus, Information
> School of the University of Washington
> Email: [log in to unmask] <mailto:[log in to unmask]>
> Skype: sasutton
>
>
|