RE: PROV-ISSUE-8: defining generation in terms of `IVPT of'

I think Stian’s arguments are along the lines of Galton and Mizoguchi’s idea I mentioned in the previous post – our notion of objects/things is tied to the set of processes we care about and hence objects are views too (IVPTs mutable with respect to processes considered part of the object’s lifecycle and invariant to (creatable/destroyable by) processes that are not). In practice, we all mostly care about the same set of processes, so ‘egg’ makes sense. But when we try to record provenance, we see that our real definitions of egg differ by just a process or two – does heating change it, does taking the shell off? Given that there are lots of potential processes, there are lots of potential IVPTs, but again, in practice, a lot of use cases will look like two (resource/state) or three (FRBR work/expression/manifestation). When those limited levels don’t work, we hop back to realizing that the invariant ‘state’ really is an IVPT that we can nail down further (we can push more processes out of its lifecycle) and that the mutable resource is just a view on something even more mutable (e.g. a bag of molecules, just atoms, a whole bunch of quantum wave functions, vibrating strings, …).

 

Stian makes another good point – when we select the set of processes we care about, we also may shift the physical definition a bit – an egg, or an egg in a shell if we consider processes that can remove the shell. I think that means that an IVPT relationship is not just 1-1 – the ‘boiled egg’ as delivered by the waiter and the combined (‘edible egg’ + ‘shell fragments’) are two IVPTs of the more general ‘egg’. 

 

Jim

 

From: public-prov-wg-request@w3.org [mailto:public-prov-wg-request@w3.org] On Behalf Of Luc Moreau
Sent: Friday, June 10, 2011 9:05 AM
To: Stian Soiland-Reyes
Cc: public-prov-wg@w3c.org
Subject: Re: PROV-ISSUE-8: defining generation in terms of `IVPT of'

 

Stian,
I agree with your comments, except for one thing:

 The asserter has made *assertions*. The asserter may or may not have observed the events; it may be hearsay, it may be inference.

But otherwise, I thought that the idea of perspective was capturing what you describe.  Doesn't it?

Cheers,
Luc



On 06/10/2011 01:26 PM, Stian Soiland-Reyes wrote: 

This seems to bring up again the notion of an Observer. The asserter has made *observations* of the egg going in and out of the hot
water, and knows about the *process* of "boiling an egg" and his intentions of eating it.

Another (occasional) observer might only have observed the egg being taken out of the fridge and it put in cold water, and later that it was on a plate. He might assert that the egg has been in water, and might suggest the process is just rinsing.

(I'm not suggesting scientists don't know what procedure they are using, just that it might be loosely or wrongly defined) 

I think this touches on Identify as well. Your observations and provenance assertions are affected by your views of "what the thing is". If your view is that "the egg" is food, you care about how it was prepared and perhaps if the hen was free ranging. Your "egg identity" includes the egg as it appears cooked on the plate without egg shell. 

As a biologist you care about how the egg was formed and play a role in reproduction, and would consider boiling and removing the shell a destructive process yielding new parts which are no longer "an egg".

So you might both make  observations and form a view of "the same egg", but only for a restricted period of time. Remember that any observation spans an interval, say the exposure time of a camera, while the view spans a longer interval because of assumptions about the object being in an expected state before, during and after the observations. (intervals might have.varying accuracy)

Later observations might contradict your earlier assumptions, of course. It might be discovered that the yolk was solid and eggwhite running, suggesting the egg had been exposed by microwaves.

(I propose enough egg analogies for now!)

On Jun 10, 2011 9:55 AM, "Luc Moreau" <L.Moreau@ecs.soton.ac.uk> wrote:





-- 
Professor Luc Moreau               
Electronics and Computer Science   tel:   +44 23 8059 4487         
University of Southampton          fax:   +44 23 8059 2865         
Southampton SO17 1BJ               email: l.moreau@ecs.soton.ac.uk  
United Kingdom                     http://www.ecs.soton.ac.uk/~lavm