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

Re: Some thoughts about the revised provenance Model document

From: Graham Klyne <graham.klyne@zoo.ox.ac.uk>
Date: Fri, 21 Oct 2011 17:20:55 +0100
Message-ID: <4EA19BE7.80608@zoo.ox.ac.uk>
To: public-prov-wg@w3.org
On 21/10/2011 15:16, Stian Soiland-Reyes wrote:
> On Fri, Oct 21, 2011 at 10:34, Graham Klyne<Graham.Klyne@zoo.ox.ac.uk>  wrote:
>
>
>> Rather than "lifetime of the account", why not "Lifetime of the entity
>> described."  Unless I've missed something, this discussion is entirely about
>> statements made directly or indirectly about a prov:Entity, which by it's
>> nature is completely static or somehow constrained in its existence.
>
> Because you might not be able to group non-provenance assertions by
> entities - several prov:Entity instances might have been located in
> the same (RDF-wise) Berlin - and then anything said about Berlin
> should be true for all such entities.

Er... if they're non-provenance assertions, does it matter?  Stuff which is said 
about Berlin may be contextualized in implicit ways, hence not necessarily true 
for all Entities that are particular instances of Berlin.

> Also an Entity could outlive an account - for instance Berlin and
> Klaus still exists, and so does his paperweight described by :e2 - but
> would :e2 no longer exist if Klaus ceases to be mayor of Berlin?
> (Perhaps that is the definition of :e2 - or it is just an incidental
> fact) When comparing two different accounts with different timespans
> these kind of questions would come up.

I'm struggling to follow this, but I'm not convinced.  I think it would help to 
try and make a real RDF example out of this dealing with the particular concerns 
raised here, and see if it can be made to work.

> I believe Jim's proposal is easier to deal with for someone producing
> the account, when deciding if some information should be included or
> not, but does raise the question of "what is the time period the
> account is valid for" - this is not necessarily continuous, but must
> cover all the events that have been described, such as
> wasGeneratedBy(), used(), etc.

Part of my concern is that I feel it should be possible to discuss provenance 
without getting caught up in accounts.  From my reading of OPM, accounts are 
quite a subtle notion that would not always be relevant at the level of saying 
(say) document X was derived from data Y.

#g
--
Received on Friday, 21 October 2011 16:21:53 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Thursday, 26 April 2012 13:06:46 GMT