RE: Provenance Working Group resolution ISSUE-447 and ISSUE-500 (subactivity)

Hi Paul,

I know this is a tricky issue.  Thanks to you and the work group for considering it (again).

I think the proposed solution (and pending FAQ/example) may provide the ability to represent subactivities, at least for now.  It will be interesting to see what use cases arise when PROV is put into practice more widely.  The example that I provided (http://lists.w3.org/Archives/Public/public-prov-wg/2012Sep/0242.html) is obviously a hypothetical one, but one that I believe PROV should support.  As I stated in that thread, I believe capturing the relationship between activities is just as important as capturing the relationship between entities.

Thanks,
Bob


________________________________
From: satrajit.ghosh@gmail.com [mailto:satrajit.ghosh@gmail.com] On Behalf Of Satrajit Ghosh
Sent: Thursday, September 27, 2012 4:02 PM
To: Paul Groth
Cc: Freimuth, Robert, Ph.D.; public-prov-comments@w3.org
Subject: Re: Provenance Working Group resolution ISSUE-447 and ISSUE-500 (subactivity)

dear paul,

thank you for the update.

ISSUE-447 (subactivity)

Original email:
http://lists.w3.org/Archives/Public/public-prov-comments/2012Jul/0003.html

Tracker: http://www.w3.org/2011/prov/track/issues/447

Group Response

- The Working Group charter identified an initial set of concepts, and
made it clear that the working group should not delve into the details
of plans and workflows (called then recipe). The charter did not list
a notion of subactivity either.

i understand trying to stay away from plans and workflows and possibly not relive the uml discussions. however, even in a simple context activities are typically related to each other in a provenance sense, and while time covers some aspect of that, it doesn't in anyway cover sub-activities.

- The Working Group considered a notion of subactivity, but does not
understand the implication of introducing such a relation to the
model. In fact, there is little prior art about this in the provenance
community. There is also concern that specifying such a relation would
overlap with some workflow specification initiatives.

that's what i was hoping a simple relation such as wasRelatedTo(a1, a2, --) would cover this and one that could then be decorated by dcterms:hasPart, partOf, etc.,.

also i would love to know about the workflow specification initiatives. as an architect of a workflow framework for brain imaging, standardizing that effort would be quite useful.

- For this reason, the Working Group decided not to provide a
normative definition of such a relation. Instead, the Working Group
suggests that a relation such as dcterms:hadPart could used by
applications, which would be responsible for ensuring its use is
consistent with the model.

- The Working Group intends to produce an FAQ page illustrating how
such a construct could be used.

really looking forward to this faq, especially where it can capture such relations as partOf.

cheers,

satra