W3C home > Mailing lists > Public > public-xg-prov@w3.org > February 2010

Re: flagship use cases proposal

From: Jim McCusker <mccusker@gmail.com>
Date: Fri, 19 Feb 2010 13:49:12 -0500
Message-ID: <68084f3e1002191049t33ae02cbt6ea4950294f41f71@mail.gmail.com>
To: Paul Groth <pgroth@gmail.com>
Cc: James Cheney <jcheney@inf.ed.ac.uk>, Yolanda Gil <gil@isi.edu>, "<public-xg-prov@w3.org>" <public-xg-prov@w3.org>
>From a personal interest, I would vote for bioinformatics. I have been
pushing adoption of a generalized provenance model for interchange in NCI's
caBIG, and the use case below is core to research in that domain.


On Fri, Feb 19, 2010 at 12:24 PM, Paul Groth <pgroth@gmail.com> wrote:

> Hi James and All,
> It seems that we have agreement on a use case e.g. a scientist uses linked
> data, processes it with a scientific workflow + some manual and qualitative
> analysis makes it available.
> The question seems which domain: eGovernment Public Policy or
> bioinformatics. There are benefits to both.
> * eGovernment has the whole push with Data.gov.uk and open government
> data, which has been really a hit with the community as a whole.
> Non-scientists can also usually understand policy type use cases.
> * for bioinformatics it would cement our ties with the HCLS working group.
> I know there are strong demands for provenance and several iniatives their
> trying to capture provenance type information. Also workflows and linked
> data have fairly strong user communities in the domain.
> I think the best way to solve this is who takes initiative :-)
> So is there anyone who would like to write up this use case (use case #2)?
> Thanks!
> Paul
> James Cheney wrote:
>> On Feb 18, 2010, at 8:26 PM, Yolanda Gil wrote:
>>  Hi Paul:
>>> Your proposal is very reasonable, thanks as always for pushing the
>>> requirements document forward and doing this synthesis.
>>> I wonder what others think, but one thing I'd suggest to keep in mind is
>>> who our main immediate consumers are going to be.  Right now my sense is
>>> that it is going to be the Linked Data community, and if I am right it is a
>>> pity that there is no use case centered around that topic.  If you want to
>>> keep the number to 3, I would say that if I had to take one of your proposed
>>> 3 out it would be #2 (public policy), as that area is kind of behind in the
>>> times for social reasons.  Just my 2c though, what do others think?
>> Hi Yolanda,
>> I'm not sure I understand what you mean by "behind the times".  It is true
>> that the use case is written more from the perspective of the traditional
>> pen-and-paper approach to research but the social scientists involved in
>> this kind of work are very interested in moving parts of it online and
>> available in ways that provide greater value to funders and society.  This
>> kind of data would, I think, be a big part of the eGovernment visions
>> discussed last week.
>> I view linked data (or curated databases, which seem like the same thing
>> to me) as very relevant to the evidence/public policy use case since online,
>> shared and linked data is likely the future of data management for these
>> areas (as they are already central in bioinformatics).
>> Likewise, social science research tends to involve a mix of qualitative,
>> manual steps and quantitative, computational steps.  So, to address Luc's
>> concern, the scenario I described could also highlight provenance in
>> scientific workflows, as well as the issue of integrating linked/curated
>> data provenance and workflow provenance.
>> So it would make more sense to me to broaden the evidence use case to
>> clarify its relevance to linked data and scientific workflow settings,
>> rather than simply replace it.
>> Of course, since I wrote the use case, naturally I'm partial to it :)
>> --James

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

PhD Student
Tetherless World Constellation
Rensselaer Polytechnic Institute
Received on Friday, 19 February 2010 18:50:00 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 19:55:59 UTC