Re: Renaming of ProcessExecution to Activity

On Fri, Dec 2, 2011 at 10:21 AM, Stian Soiland-Reyes <
soiland-reyes@cs.manchester.ac.uk> wrote:

> Perhaps the two of us can accept Activity for now, but have in the
> back of our mind to try to think of any good "no agent, not an
> activity" examples which would easily be described with PROV. I'm not
> sure if the provenance of a glacier is the best fit..  :)
>
> I mainly objected because of the preferred style of using
> PEs/activities to model "state transitions" between static entities
> which are representing "the same thing in the world". If
> :roofWithWater is derived from :dryRoof and :rain - then there was now
> an "activity" called Raining which caused the transition - but who
> performs the Raining activity? The rain? The weather? The cloud? The
> roof?
>

That should be describable using PROV.

One example that is useful to think about is a genetic disorder like
Huntington's Disease. HD is a mutation in the gene that codes for
huntingtin, and is deterministic: it will show up when it's going to, and
won't really be affected by environment or behavior. We should be able to
say things like:

activity(HDOnsetInJohnDoe,HuntingtonsDisease,12/1/2011)
entity(JohnDoeBeforeHDOnset)
used(HDOnsetInJohnDoe,JohnDoeBeforeHDOnset)
entity(JohnDoeBeforeHDOnset,[hasSymptom=Chorea])
wasGeneratedBy(JohnDoeAfterHDOnset,HDOnsetInJohnDoe)

No where in this event is an agent. Calling it an activity is very awkward
because of this, IMO.

Jim
-- 
Jim McCusker
Programmer Analyst
Krauthammer Lab, Pathology Informatics
Yale School of Medicine
james.mccusker@yale.edu | (203) 785-6330
http://krauthammerlab.med.yale.edu

PhD Student
Tetherless World Constellation
Rensselaer Polytechnic Institute
mccusj@cs.rpi.edu
http://tw.rpi.edu