W3C home > Mailing lists > Public > public-prov-wg@w3.org > July 2011

Re: PROV-ISSUE-42 (derivation-agent): Derivation should specifically mention agent in its definition [Conceptual Model]

From: Paul Groth <p.t.groth@vu.nl>
Date: Mon, 25 Jul 2011 13:25:50 +0200
Message-ID: <4E2D52BE.3090706@vu.nl>
To: martin <martin@ics.forth.gr>
CC: "public-prov-wg@w3.org" <public-prov-wg@w3.org>
Dear Martin,

I think this notion of transferring relevant features of a source is in 
the definition. There is the following sentence as part of the definition:

"From an assertion*isDerivedFrom(B,A)*, the values of some 
characteristics of B are at least partially determined by the values of 
some characteristics of A. "

which I think reflects that notion... and the constraint may need to be 
tightened.

Thanks,
Paul

martin wrote:
> Dear All,
>
> I may miss here something, but if derivation is nothing else than the closure of all used things,
> I do not see any point in calling it "derivation", but right away "use". This definition extends the
> intuition of derivation beyond recognition. For instance, it would include all raw materials.
> Also, I do not understand how a "characterized entity" could "affect" another, if it is not an agent.
>
> I'd prefer a definition of derivation based on a notion that relevant features of a source are transferred
> to the "derivative". This makes also sense for IPR.
>
> Best,
>
> Martin
>
> On 7/25/2011 12:54 PM, Paul Groth wrote:
>> Hi Luc,
>>
>> looking at the definition of Derivation it says:
>>
>> "Derivation expresses that some characterized entity is transformed from, created from, or affected by another characterized entity."
>>
>> I think I'm thinking of the "created from" part of the definition in my example. I want to say explicitly that David (e0) created an
>> article(e1). Notationally: isCreatedFrom(e1, e0)
>>
>> I think this is compatible with the definition as it stands but not compatible with the inference rule you propose. It would seem bizzare to
>> say that a process used a person in this example...
>>
>> Could you explain how that should be represented using the concepts we have?
>>
>> Several of the shortcuts I think we need rely on making simple statements about agents and their relationship to an entity. I thought the
>> best approach was to create specializations of isDerivedFrom but maybe that's not the best approach and it would be good to understand that
>> better.
>>
>> Thanks,
>> Paul
>>
>>
>>
>>
>> Luc Moreau wrote:
>>> Yes, I have no problem for agents to be source/destination of a
>>> derivation, but your example
>>> may introduce some confusion. Let me try and explain why.
>>>
>>>
>>>
>>> First, I think there is a missing "inference" in the specification.
>>>
>>> If there isDerivedFrom(e1,e0) holds, then there exists a process
>>> execution pe, and roles r0,r1,
>>> isGeneratedBy(e1,pe,r1) and use(pe,e0,r0)
>>>
>>>
>>> So, if I apply this to your example,
>>>
>>> isGeneratedBy(e0,pe,r1) and use(pe,David,r0)
>>>
>>>
>>> David may have been asserted to be an agent, or the agent nature of
>>> David can be inferred (as per definition
>>> of agent), but it's not because of its involvement in pe. It has to be
>>> in another process execution, right?
>>>
>>> Maybe, the example could become:
>>>
>>> isDerivedFrom(david-in-his-thirties, david-in-his-twenties).
>>>
>>> What do you think?
>>> Luc
>>>
>>>
>>>
>>> On 07/23/2011 04:36 PM, Provenance Working Group Issue Tracker wrote:
>>>> PROV-ISSUE-42 (derivation-agent): Derivation should specifically mention agent in its definition [Conceptual Model]
>>>>
>>>> http://www.w3.org/2011/prov/track/issues/42
>>>>
>>>> Raised by: Paul Groth
>>>> On product: Conceptual Model
>>>>
>>>> Given that isDerivedFrom is between Bobs this by definition allows it to relate agents, it would nice for informative to mention this in
>>>> the definition.
>>>>
>>>> For example, I would like to say that isDerivedFrom(e0, David) this is fine with the current definition but might not be clear.
>>>>
>>>> Suggested resolution:
>>>>
>>>> Add the following statement: "Note, that isDerivedFrom can also include agents. For example, isDerivedFrom(e0, David).
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>
>

-- 
Dr. Paul Groth (p.t.groth@vu.nl)
http://www.few.vu.nl/~pgroth
Assistant Professor
Knowledge Representation & Reasoning Group
Artificial Intelligence Section
Department of Computer Science
VU University Amsterdam
Received on Monday, 25 July 2011 11:28:43 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Thursday, 26 April 2012 13:06:37 GMT