Re: PROV-ISSUE-205 (optional-activity-in-generation-record): optional activity in generation record [prov-dm]

On Thu, Jan 12, 2012 at 09:44, Khalid Belhajjame
<Khalid.Belhajjame@cs.man.ac.uk> wrote:

> In addition to the qualified generation, prov-o allows defining the object
> property wasGeneratedBy between an entity and an activity. If we allow for
> the creation of qualified generation that is not associated with an
> activity, it means that in certain cases we may have qualified generation
> without having a corresponding wasGeneratedFor property. I am wondering if
> that may lead to inconsistencies at the level of the ontology.

You are right, there is nothing with QualifiedInvolvement that
requires it to be bound to an activity, but in all our uses of it
there is an implicit property of hadQualifiedActivity as the purpose
of a QI is to link an entity to an activity with some additional
properties.

The properties of Activity; hadQualifiedControl, hadQualifiedEntity
etc. can be thought of as subproperties of an imaginary
hadQualifiedInvolvement - which would have been inverse of
hadQualifiedActivity. However using hadQualifiedInvolvement alone
would also be confusing, and hadQualifiedActivity would just be a
superficial information.

I don't think we can put this constraint into the OWL file directly
without affecting the OWL-level or introducing lots of extra
properties, but it is probably a constraint we should keep in mind for
separate rules.


Tim - did we have something on this in the earlier QI proposals?

-- 
Stian Soiland-Reyes, myGrid team
School of Computer Science
The University of Manchester

Received on Thursday, 12 January 2012 14:09:11 UTC