Re: PROV-ISSUE-268 (two-level-ontology): Two Level Ontology? [Ontology]

Hi Luc,

The PROV-O team discussed the two-level-ontology idea during our Monday telecon and came to the conclusion that at the present time we are averse to developing/supporting multiple levels of prov ontologies.

The WG is averse to multiplying our responsibilities by adding another ontology and related documentation that must be maintained and released along other PROV WG deliverables.  It will be difficult to maintain a clear distinction between the ontologies without simplifying one ontology to the point where it no longer faithfully represents the PROV-DM; which would introduce incredible challenges to interoperability and likely cause confusion among users regarding which ontology the WG recommends.

The group feels that proceeding with a single ontology targeted to the 'OWL2-RL++' profile is currently the best course of action.

--Stephan


On Feb 24, 2012, at 1:45 AM, Provenance Working Group Issue Tracker wrote:

> PROV-ISSUE-268 (two-level-ontology): Two Level Ontology? [Ontology]
> 
> http://www.w3.org/2011/prov/track/issues/268
> 
> Raised by: Luc Moreau
> On product: Ontology
> 
> 
> Dear all,
> 
> For the record, I made a suggestion to Khalid yesterday, and it would be good if the prov-o team could consider it.
> 
> The details are not fully worked out, and I am sure lots of variants are possible.
> 
> The essence is to consider two separate ontologies:
> - one minimalistic, a simple vocabulary, in which we allow (more or less) the same expressivity as in PROV-DM
> - the other, more extensive, which provides a structure to the vocabulary, introduce super-classes and super-relations, has property chains, has more complex constraints.
> 
> For the purpose of this email, I call them prov and provs (for structure)
> 
> I believe this would address multiple concerns
> - ISSUE-262, ISSUE-263: some of the more permissive assertions would be in provs not in prov. For me this solves the alignment issue.
> 
> - ISSUE-265: prov only is required to be OWL-RL (I think it could even be RDFS). provs does not have to be restricted by any specific profile.
> 
> Concretely, in the email to Khalid
> http://lists.w3.org/Archives/Public/public-prov-wg/2012Feb/0413.html,
> I suggested the following
> 
> 
> :a1 a prov:Activity
>   prov:used :e1
>   prov:usage [a Usage
>                       prov:usedEntity  :e1
>                       prov:usedTime t]
> 
> 
> Then, in prov-s (s for structure)
> 
> 
>  prov:usedEntity subPropertyOf provs:entity
>  prov:Usage subclassOf provs:EntityInvolvement
>  prov:usedTime subRelationOf provs:hadTemporalExtent
>  provs:entity domain: provs:EntityInvolvement
>                      range  prov:Entity
> 
>   prov:usage subrelationOf provs:qualified
>   provs:qualified domain: provs:Element
>                            range: provs:Involvement
>   prov:Activity subclassOf provs:Element
>   prov:Entity subclassOf provs:Element
> 
> 
> 
> All the patterns are preserved. The concern about Involvement not
> being abstract has disappeared. In prov, you can't express instance
> of involvement, it's only in provs you can.
> 
> 
> 
> 
> 
> 
> 
>

Received on Wednesday, 2 May 2012 21:52:46 UTC