W3C home > Mailing lists > Public > public-prov-wg@w3.org > February 2012

Re: PROV-O ontology comments

From: Jun Zhao <jun.zhao@zoo.ox.ac.uk>
Date: Thu, 16 Feb 2012 14:39:56 +0000
Message-ID: <4F3D153C.7060000@zoo.ox.ac.uk>
To: public-prov-wg@w3.org
Dear all,

Here is my homework.

On 16/02/2012 11:18, Simon Miles wrote:
> Hello PROV-O team,
>
> Here's my quick review of the PROV-O ontology, as requested last telecon.
>
> In general, it seems fine.
>
> It seems odd that hadLocation has a domain of owl:Thing. So what is
> it's connection to provenance?

There are also some other object properties appear odd to me:

- hasQualifiedControl, there is no range definition

- hasQualifiedEntity, should the domain be some sort of subclass of 
QualifiedInvolvement?

- wasEndedBy and wasStartedBy do not have domain or range definitions


>
> What does it mean that hasAnnotation does not have a specified domain
> (my ignorance of RDFS)? If it means that it applies to anything, then
> what is the distinction between using hasAnnotation and just giving an
> arbitrary non-prov RDF statement? What is its connection to
> provenance?
>
> The old W3C 2006 Time namespace is still used/included. Is use of this
> ontology to be removed in the next revision?

Same question.
>
> I notice that comments (at least for adoptedPlan) still refer to
> ProcessExecution.

And there are 9 (!) object properties to associate an entity with 
another entity or a subclass of it. That's a lot of different types of 
object properties between two classes! Is there any way to simplify it, 
either in the DM doc or in the ontology?

There are not enough annotations in the ontology. I still need to spend 
some more time to read the DM doc to understand what classes like Plan, 
Association or Bundle are for. My bad. :(


Talk to you soon!

-- Jun


>
> Thanks,
> Simon
>
Received on Thursday, 16 February 2012 14:40:25 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Thursday, 26 April 2012 13:06:56 GMT