W3C home > Mailing lists > Public > public-xg-prov@w3.org > November 2010

Re: Grouping of concepts

From: Paulo Pinheiro da Silva <paulo@utep.edu>
Date: Fri, 26 Nov 2010 09:03:40 -0700
Message-ID: <4CEFDA5C.3020105@utep.edu>
To: Paul Groth <pgroth@gmail.com>, "public-xg-prov@w3.org" <public-xg-prov@w3.org>
Hi Paul,

Thank you a lot for grouping these concepts.

This grouping is truly interesting for me since since it shows how the 
name of certain concepts can be misleading. For example, I did mention 
that PML concepts added to the list are those that we do not have OPM 
equivalent. In this case, the PML term that maps to opm:Artifact is 
pml:Information that is not in my list. pml:Document is a subclass of 
pml:Source that is an information container. What we mean by this? Well, 
opm:Artifact is an immutable state of an object but that the object it 
is a state of is itself mutable. So, pml:Source.

So, what is the relevance of pml:Document? We have noted that objects 
can be actionable and not-actionable. More specifically in terms of 
provenance, some objects are capable of asserting information where they 
are the originators of the information (e.g., a person, an organization) 
while other objects are capable of asserting information that was once 
provided to them (e.g., a publication, a database).

As you see, that is why my list have pml:Source (and its subclasses 
pml:Agent, pml:Document and pml:DocumentFragment) and it does not have 
pml:Information. Moreover, I do not include some subclasses to this list 
to keep it small such as pml:Person and pml:Organization (subclasses of 
pml:Agent) and pml:publication (subclass of pml:Document).

Many thanks,

> Hi All,
> Thanks for all your contributions to the set of suggested concepts on
> http://www.w3.org/2005/Incubator/prov/wiki/Proposal_for_a_Working_Group_on_Provenance
> I have went ahead and tried grouping those concepts together across the
> various ontologies. You can find this grouping here:
> http://www.w3.org/2005/Incubator/prov/wiki/Suggested_Concepts
> In this grouping, I have been biased to condensing the concepts to make
> the most manageable list possible. For example, I have grouped all the
> concepts related to time under Time or for example I have grouped
> opm:artifact, provenier;data, and pmlp:document under Resource. I have
> made no distinction between classes and properties. Note, I may have put
> an ontological concept in the wrong grouping but I have not eliminated
> any concepts.
> According my grouping that we have roughly 19 core concepts.  Today, on
> the call I hope we can get final consensus on these core concepts
> whether some should be included or not.
> Any comments are appreciated. I hope this is helpful.
> Speak to you soon,
> Paul
> .
Received on Friday, 26 November 2010 16:04:15 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 19:56:00 UTC