Re: PROV-ISSUE-124: Constraints on Used Relation (PROV-DM and PROV-OM) [Conceptual Model]

Hi Satya,

Responses interleaved.

On 11/10/11 19:24, Provenance Working Group Issue Tracker wrote:
> PROV-ISSUE-124: Constraints on Used Relation (PROV-DM and PROV-OM) [Conceptual Model]
>
> http://www.w3.org/2011/prov/track/issues/124
>
> Raised by: Satya Sahoo
> On product: Conceptual Model
>
> The following constraints are defined for Used Relation (in PROV-DM document Oct 11, 2011):
>
> Constraint 1: "Given a process execution expression identified by pe, an entity expression identified by e, a qualifier q, and optional time t, if assertion used(pe,e,q) or used(pe,e,q,t) holds, then the existence of an attribute-value pair in the entity expression identified by e is a pre-condition for the termination of the activity represented by the process execution expression identified by pe."
>
>    

Can you use the constraint ids, so that we can find them easily in the 
document? thanks.
> Issue:
> a) The above constraint may not hold for many scenarios involving Used relation. For example, if "table salt" was added by mistake to a cakeBaking PE, then Used(salt, cakeBaking PE) is true, but it is not clear what attribute-value must exist for "salt" to allow cakeBaking PE to terminate?
>    

Khalid had suggested that

- the existence of an attribute-value pair in the entity expression identified by e
or
- the very existence of the entity expression
... are pre-conditions


This latter clause seems to have been lost. It may address your concern.


> b) Without specifying the identity, the characteristics, and how does this "attribute-value pair" relate to the Entity e itself (is it a necessary attribute-value pair for existence of e etc.) it is unclear how can we use this constraint.
>    

As a group, we need to decide whether we want the constraints to help 
define the interpretation of the data model,
or we want them to validate assertions (a comment you have made on 
several occasions).

My view (and Paolo noted that this was not explicit in the document) is 
that some constraints are just here to help
provide interpretation to the data model.

We may want to opt for the second option, but this would require a lot 
more to be exposed in the data model.
For instance, here, we would have to identify which attribute is 
pre-condition to the end.

I think we need to have a debate about what we want to express.

> c) Further, is it necessary for the attribute-value to be explicitly stated prior to the start of PE instance - since with the open world assumption it may exist but not known to a provenance application before start of PE.
>    

Do you mean asserted before the start of PE?  I don't understand.

Use relation has nothing to say about its order with respect to the 
start of the process execution.

> --------------
>
> Constraint 2: "Given a process execution expression identified by pe, an entity expression identified by e, a qualifier q, and optional time t, if assertion used(pe,e,q) or used(pe,e,q,t) holds, then the use of the thing represented by entity expression identified by e precedes the end time contained in the process execution expression identified by pe and follows its beginning. Furthermore, the generation of the thing denoted by entity expression identified by e always precedes its use."
>
> Issue:
> To enforce this constraint, it will be necessary for "time" (or events?) to be associated with both PE and Entity instances to derive ordering - currently association of time is optional for both PE and Entity (events is not defined).
>    

See my answer to b).  The intent of this constraint was just to give an 
interpretation to the data model.

If we want to be able to verify it, then, yes, it looks that we would 
have to make explicit some events.

Let's debate our intent. We adjust the model, if required, depending on 
the outcome.

Luc
>
>
>

Received on Tuesday, 11 October 2011 21:14:00 UTC