Re: PROV-ISSUE-121: Constraint on PE (PROV-DM and PROV-OM) [Conceptual Model]

Hi Satya,

The ordering constraints are all now in a single section
http://dvcs.w3.org/hg/prov/raw-file/default/model/ProvenanceModel.html#interpretation

ISSUE-82 was closed after a month without any objection to its proposed 
resolution.

I am proposing to close this issue, pending review.

Best regards,
Luc



On 10/09/2011 10:54 PM, Provenance Working Group Issue Tracker wrote:
> PROV-ISSUE-121: Constraint on PE (PROV-DM and PROV-OM) [Conceptual Model]
>
> http://www.w3.org/2011/prov/track/issues/121
>
> Raised by: Satya Sahoo
> On product: Conceptual Model
>
> PE Constraint defined in the PROV-DM document (as on Oct 9, 2011):
> "The mere existence of a process execution assertion entails some event ordering in the world, since the start event precedes the end event. This is expressed by constraint start-precedes-end.
> > From a process execution expression, one can infer that the start event precedes the end event of the represented activity."
>
> There are multiple issues with the above constraint:
>
> 1. The constraint is defined with respect to events (in previous version of PROV-DM it was defined with respect to time), and event (a) is not defined, and (b) is not part of either the definition of PE or the PE expression. Hence, it is not clear how can this constraint can be defined and enforced for PE?
>
> In other words, the "mere existence of a process execution" cannot entail "some event ordering in the world" since a PE can be defined without making any assertion about events (start or end).
>
> 2. Issue 82 discussed the introduction of event as concept but there was no final decision, hence pending clarification about its status in PROV (both DM and OM), we should not use it for defining constraint that need to be satisfied by provenance applications. Further, given the current use of time with PE definition and PE expression, it is more intuitive to state the constraint as:
>
> "The start time of a PE instance precedes the end time of a PE instance" where time measurement is application-specific requirement
>
> 3. Assuming, we go back to the original formulation of this constraint in terms of time - If this is asserted as a constraint to be satisfied by PROV compliant provenance applications, it will make association of time value with PE a necessary condition and not optional (as the current definition of PE states).
>
>
>
>    

-- 
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

Received on Tuesday, 20 December 2011 16:38:42 UTC