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

Re: Different senses of specialization

From: James Cheney <jcheney@inf.ed.ac.uk>
Date: Mon, 9 Apr 2012 23:35:50 +0100
Cc: <public-prov-wg@w3.org>
Message-Id: <50E106F5-48E6-4DD7-9C9F-FEB4D099A44B@inf.ed.ac.uk>
To: Curt Tilmes <Curt.Tilmes@nasa.gov>

On Apr 9, 2012, at 5:19 PM, Curt Tilmes wrote:

> On 04/09/2012 10:57 AM, James Cheney wrote:
>> To summarize:
>> - I think we should be careful about these different senses of "specializationOf".
>> - Only the first sense is supported by the current version of the semantics.
>> - If we can agree on one of these definitions for specializationOf, but believe other senses need to be modeled, we should introduce additional relations to name them, and ensure that the meanings are clear and they are used consistently in examples.
> I'm trying to relate my example:
> http://lists.w3.org/Archives/Public/public-prov-wg/2012Apr/0027.html
> with your descriptions.
> I think there is a need for a FRBR like system for the types of
> representations I am trying to express.  I suspect it is too
> complicated for PROV to attack the whole problem right now.
> One of the concepts we've been working with is relating two things
> with a concept I've called "provenance equivalence" i.e. two things
> that were made in essentially[1] the same way.
> [1] essentially = Those elements of provenance from which their
> essence is derived.
> Building on my previous example:
> Suppose I don't archive the data at all, I have a process on demand
> system that generates a data file whenever it is requested.
> User A gets file x and user B gets file y.
> They are made by the same system in the same way, using a process that
> is supposed to be reproducible.  The intent of the system developer is
> that x and y be the same (be equivalent).
> file x and file y are different.  They were made at different times,
> have different agents that start them, etc. but they also 'use' the
> same inputs and apply the same software algorithm/software, through
> very similar activities.
> What is the relationship between file x and file y?
> Is x alternateOf y?
> I would say no, based on our definitions, since we might be wrong --
> the process might not be reproducible (for whatever reason) and x and
> y might have real essential differences.

Interesting example!

I don't think alternateOf means "has the same/similar provenance" either.

I'm also not sure how you would define "hasSameProvenanceAs" - to do so, you may need to scope how far back you look and decide what parts of the provenance are "important".  For example if I make two photocopies of something on two different sheets of paper, one from Europe and the other from the US, they have the same provenance if we only look back to the photocopy step but different if we keep going.

> Even if alternateOf is the wrong thing to use for this example, it
> seems like there is a real relationship between x and y that we should
> be able to claim/assert/represent.

I agree that it is an interesting notion, but it seems out of scope of PROV to me, as nothing like it has been discussed so far.  However, nothing stops an application from allowing such an assertion, or defining different application-specific notions of "same/similar provenance".  So I am not sure I agree that PROV should include such a construct.  Obviously that is just my opinion.

The University of Edinburgh is a charitable body, registered in
Scotland, with registration number SC005336.
Received on Monday, 9 April 2012 22:36:39 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 16:51:11 UTC