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

Re: PROV-ISSUE-491: [external] feedback on prov:agent explanation. [PROV-O HTML]

From: Stian Soiland-Reyes <soiland-reyes@cs.manchester.ac.uk>
Date: Thu, 27 Sep 2012 15:56:51 +0100
Message-ID: <CAPRnXtkND9ka+XSZKaH9Smh2PnxBOeAyp1waDS7yV1nfpeRsTQ@mail.gmail.com>
To: Provenance Working Group <public-prov-wg@w3.org>
Suggested replacement definition:

> The property prov:agent describes an prov:Agent which influenced a resource. This property applies to an prov:AgentInfluence, which is given by prov:qualifiedInfluence or its subproperties from the influenced prov:Entity, prov:Activity or prov:Agent.


Tim - if I add this as a <editorsDefinition> to #agent - should I
remove the  <sharesDefinitionWith
rdf:resource="http://www.w3.org/ns/prov#AgentInfluence"/> ?


If style is acceptable, I'll replicate it for prov:entity and
prov:activity, as I agree with Patrice that just copying the
*Influence definition makes it confusing.


On Tue, Sep 4, 2012 at 3:32 PM, Provenance Working Group Issue Tracker
<sysbot+tracker@w3.org> wrote:
> PROV-ISSUE-491: [external] feedback on prov:agent explanation. [PROV-O HTML]
>
> http://www.w3.org/2011/prov/track/issues/491
>
> Raised by: Timothy Lebo
> On product: PROV-O HTML
>
> I was looking over the PROV-O writeup, and I want to raise two points.
> Pleas forward appropriately, if you deem it appropriate.
>
> If you search for "Property: prov:agent op" at
> http://www.w3.org/TR/prov-o/ you find:
>
> "The property used by a prov:AgentInfluence to cite the Agent that
> influenced an Entity, Activity, or Agent."
> 1) A property is "used" by an agent influence to "cite"--is the agent
> influence writing a specification?
>
> Perhaps: "If x is in the prov:agent relationships with y, x is an
> instance of AgentInfluence and y is an instance of Agent that
> influenced an Entity, Activity, or Agent."
>
> OR (either way should be based on an established and clear convention
> for the document)
>
> "An agent influence is in the prov:agent relationship with an agent
> only if the agent influenced an Entity, Activity, or Agent."
>
> 2)For the first sentence the passage, that as the reader I assume is
> to begin defining the prov:activity property, does not mention
> prov:agent, it should. The passage that I mention in #1 is not in the
> first passage, and in fact is after the example that is given. This is
> confusing for the reader. If you are introducing and defining a
> property, then define the property first. I see the same thing for
> prov:activity. In searching for "Class: prov:AgentInfluence" I see
> that the initial passage for prov:AgentInfluence is the same initial
> passage for prov:agent and prov:activity. So perhaps this is part of
> some auto formatting/convention?
>
> Cheers,
> Patrice
>
>
>



-- 
Stian Soiland-Reyes, myGrid team
School of Computer Science
The University of Manchester
Received on Thursday, 27 September 2012 14:57:43 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Thursday, 27 September 2012 14:57:44 GMT