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

Re: PROV-ISSUE-262 (qualifier-property): entity used entity? [Ontology]

From: Daniel Garijo <dgarijo@delicias.dia.fi.upm.es>
Date: Thu, 23 Feb 2012 14:34:57 +0100
Message-ID: <CAExK0DfBTjKkQYdO7yjAQND0u9E=ZYLT-DxNmwA7sp1H9SbZ=A@mail.gmail.com>
To: Provenance Working Group <public-prov-wg@w3.org>
Yes, I raised this on the mailing list the other day as well.
If we delete the subclasses of qualified, then we allow this kind of
assertions. On the other hand,
if we add subclasses for every involvement the model gets more complicated.
If we finally decide to add the subclasses, I wouldn't name properties the
same way as their classes
(as in usage and Usage), because it gets confusing. I like the "hadUsage",
but Tim wasn't very happy
about it.

Daniel

2012/2/22 Provenance Working Group Issue Tracker <sysbot+tracker@w3.org>

> PROV-ISSUE-262 (qualifier-property): entity used entity?  [Ontology]
>
> http://www.w3.org/2011/prov/track/issues/262
>
> Raised by: Luc Moreau
> On product: Ontology
>
>
> I am not sure, but it looks like we can write
>
> :e2 prov:qualified [a Usage
>                             prov:entity :e1]
>
> and
>
> :a2 prov:qualified [a Usage
>                             prov:entity :e1]
>
> where e2 and a2 are entity and activity respectively.
>
> It seems that the domain of qualified is any element, so it can beany
>  activity or entity.  This allows any Involvement to be linked with any
> Element, which is contrary to the table at the beginning of section 4.2 in
> the prov-dm draft.
>
> Over the week-end there was a discussion of simplifying hadQualifiedUsage
> property into something more readable.  My suggestion was hadUsage, or even
> usage. This would avoid this type of problem.
>
>
>
>
>
>
>
>
Received on Thursday, 23 February 2012 13:35:28 GMT

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