W3C home > Mailing lists > Public > public-prov-wg@w3.org > September 2011

Re: PROV-ISSUE-99: prov:eventuallyUsed - a transitive version of prov:used. [Formal Model]

From: Luc Moreau <L.Moreau@ecs.soton.ac.uk>
Date: Wed, 21 Sep 2011 20:42:57 +0000
To: Daniel Garijo <dgarijo@delicias.dia.fi.upm.es>
CC: Khalid Belhajjame <Khalid.Belhajjame@cs.man.ac.uk>, "public-prov-wg@w3.org" <public-prov-wg@w3.org>
Message-ID: <EMEW3|f390764c91f460815b536ef2b47fbba7n8KLhK08L.Moreau|ecs.soton.ac.uk|F4BF85B8-4C44-4EEE-919A-A0412E5E41D8@ecs.soton.ac.uk>
Hi daniel,

By wasTriggeredBy, I assume you mean wasInformrdBy.  It is not a transitive property and I realise the spec does not say it.

Only entities can be generated by PEs. So again, there seems to be a type error.

Professor Luc Moreau
Electronics and Computer Science
University of Southampton
Southampton SO17 1BJ
United Kingdom

On 21 Sep 2011, at 18:54, "Daniel Garijo" <dgarijo@delicias.dia.fi.upm.es<mailto:dgarijo@delicias.dia.fi.upm.es>> wrote:

I like the generation + dependedOn too. We should start discussinng how to add it
to the ontology as rules, if everybody agrees.

Just to clarify further the Khalid's example, I think that the proposed change is aimed to
to model the wasTriggeredBy property between processes: if a process triggers(generates) another one
that triggers(generates) another itself, then we could say that the last one "wasEventuallyGeneratedBy" the first one.


2011/9/21 Khalid Belhajjame <<mailto:Khalid.Belhajjame@cs.man.ac.uk>Khalid.Belhajjame@cs.man.ac.uk<mailto:Khalid.Belhajjame@cs.man.ac.uk>>


The issue raised by Tim was a follow up of few emails that were exchanged between the members of the formal model sub-group.

To illustrate the transitivity that is meant, consider an entity e0 that is used by a process execution pe1, and consider that pe1 was used (consumed) by a second process execution pe2. pe1 is, therefore, both a process execution and an entity.

The question raised by the issue is whether we can state that pe2 used e0? Tim was suggesting the use of "eventuallyUsed" instead of used in this context.

The same above observation applies to generation.

Note that in the above it is assumed that the classes prov:Entity and prov:ProcessExecution are not necessarily disjoint.

Now, in your answer, you were suggesting that:
 used(pe,e) or  (used(pe,e1) and dependedOn(e1,e)) implies eventuallyUsed(pe,e).

I quite like this. And I think we can extend it for the case of generation. That is:

wasGeneratedBy(e,pe) or (wasGeneratedBy(e1,pe) and dependedOn(e,e1)) implies wasEventuallyGeneratedBy(e,pe)

Thanks, khalid

On 19/09/2011 20:38, Luc Moreau wrote:
Hi Tim,

What do you mean by transitive here, given that domain  is ProcessExecution and
range Entity?

Maybe, you would like to define
eventuallyUsed(pe,e) if
 used(pe,e1) and dependedOn(e1,e)


On 19/09/11 19:35, Provenance Working Group Issue Tracker wrote:
PROV-ISSUE-99: prov:eventuallyUsed - a transitive version of prov:used. [Formal Model]


Raised by: Timothy Lebo
On product: Formal Model

To clarify whether used is transitive or not, I propose a transitive superproperty "eventuallyUsed" to make the distinction clear.

The corresponding considerations also need to be made for the conceptual document.

The OWL axioms related to this property are at

   a owl:ObjectProperty, owl:TransitiveProperty;

prov:used rdfs:subPropertyOf prov:eventuallyUsed .
Received on Wednesday, 21 September 2011 20:44:27 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Thursday, 26 April 2012 13:06:41 GMT