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

Re: Best practice for specialization

From: Jim McCusker <mccusj@rpi.edu>
Date: Sat, 31 Mar 2012 12:28:16 -0400
Message-ID: <CAAtgn=TDJcHm6EURP2BdFsobG+vYZAW1MgKPjtACifH_tCx-bw@mail.gmail.com>
To: "Miles, Simon" <simon.miles@kcl.ac.uk>
Cc: Provenance Working Group <public-prov-wg@w3.org>
I would say that the best practice would be to try to conform to the FRBR
model. We (RPI) have outlined application of FRBR to information resources (
http://tw.rpi.edu/web/doc/parallelIdentitiesOGD) and have produced a
mapping of FRBR to PROV (FRIR, in submission at IPAW). The OWL for FRIR (
http://purl.org/twc/ontology/frir.owl) uses the vocab.org version of FRBR (
http://vocab.org/frbr/core).

To answer your question more directly, option 2 seems to be the safer
approach. You should probably redirect :docInGeneral to :docv2 if you're
using this for linked data.

In FRBR, docInGeneral is a Work and docv1 and doc are both expressions.

Jim

On Sat, Mar 31, 2012 at 12:07 PM, Miles, Simon <simon.miles@kcl.ac.uk>wrote:

>   Hello,
>
> In working on the primer examples, I thought about a situation which it
> seems could commonly occur, but for which the best solution is not clear to
> me.
>
> I create a document and give it an ID, :doc. I make the document public
> along with metadata including PROV statements that it was generated by some
> activity and suchlike.
>   :doc a prov:Entity.
>   :doc prov:wasGeneratedBy :editing.
>   :doc dcterms:title "What I did on my holidays"
>   etc.
>
> I was not expecting to change the document, but later notice a serious
> error which means I need to correct it. I describe the new version as a new
> entity, :docV2. In describing :doc above, I was describing the document in
> general, not a specific version of it, as I didn't expect multiple versions
> to exist. I would like to say:
>   :docV2 prov:specializationOf :doc.
>
> but some of my statements above, like the wasGeneratedBy, are specific to
> the original version.
>
> So, I could either
>  (1) Create an entity :docV1, state that :docV1 and :docV2 are
> specializations of :doc, and change some of the original statements to be
> about :docV1 where specific to the original version.
>
>  (2) Don't change the original statements, treat :doc as referring to the
> original version, create an entity :docInGeneral, which :doc and :docV2 are
> specializations of, and re-assert each of the above statements about :doc
> that are also true for :docInGeneral
>
> Option (1) seems only possible if we assume copies of the statements do
> not exist elsewhere, e.g. have already been downloaded.
>
> Would guidance on this kind of situation be helpful to have in the best
> practices document?
>
> Is there more general semantic web guidance on what to do when you thought
> you were identifying one thing and it turned out to be two?
>
> Thanks,
> Simon
>
>  Dr Simon Miles
> Senior Lecturer, Department of Informatics
> Kings College London, WC2R 2LS, UK
> +44 (0)20 7848 1166
>
> Electronically querying for the provenance of entities:
> http://eprints.dcs.kcl.ac.uk/61/
>



-- 
Jim McCusker
Programmer Analyst
Krauthammer Lab, Pathology Informatics
Yale School of Medicine
james.mccusker@yale.edu | (203) 785-6330
http://krauthammerlab.med.yale.edu

PhD Student
Tetherless World Constellation
Rensselaer Polytechnic Institute
mccusj@cs.rpi.edu
http://tw.rpi.edu
Received on Saturday, 31 March 2012 16:29:07 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Thursday, 26 April 2012 13:06:59 GMT