Re: PROV-ISSUE-4: Defining Agent using FOAF's definition

On Tue, Jun 21, 2011 at 06:28, Luc Moreau <L.Moreau@ecs.soton.ac.uk> wrote:

> Reiterating a previous comment I made, can an Agent be defined independently
> of process execution?

There might be agents who DIDN'T initiate/control a process when
perhaps they should have. I'm not sure how that could be captured in
provenance - perhaps they were involved in the overall process.

In terms of example, imagine the journalist example
http://www.w3.org/2011/prov/wiki/ProvenanceExample:

* analyst (alice) downloads a turtle serialization (lcp1) of the
resource (r1) from government portal
* analyst (alice) generates a chart (c1) from the turtle (lcp1) using
some software (tools1) with statistical assumptions (stats1

Alice the Analyst agent does however *not* control or initiate a
process for Verifying the conversion data (d1) to RDF (f1). She does
not even look at the raw data d1 (perhaps it's not been published in
raw format).

If that had been in the original provenance trail then Bob the Blogger
might more easily conclude that the re-published data has been
tampered with to deal with the news story, and that the government's
argument about "something went wrong going to RDF" could be a
cover-up.


A question is if Ed the Editor should have spotted this - how was he
involved in reviewing the story and Alice's acting as an agent before
it was published? He is the missing agent. (This sounds like a poor
movie plot line)

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

Received on Friday, 24 June 2011 08:15:53 UTC