Re: Where should "Provenance" go?

Happy to see a discussion happening!

2011/9/27 Christophe Guéret <c.d.m.gueret@vu.nl>:
> On Mon 26 Sep 2011 22:04:07 CEST, John Erickson wrote:
>> Grouping 'Provenance" with "Versioning" probably makes the most sense as long as we agree that they are not the same thing ;)
>
> They are indeed two different, but not unrelated, things and I would argue it's a reason for not putting them together. Why not just keeping provenance into the vocabulary discussion? Seemed to be a reasonable setup.

Because our mission is to provide "best practices" guidance, I would
argue there are both vocabulary "policy" components
* Vocabulary, for obvious reasons: the options available for binding
provenance data to GLD
* Policy (for lack of a better term) that guides stakeholders in
making decisions about what provenance is, its implications, how to
generate it

> During our last call "pragmatic provenance" was mentionned as, if I remember correctly, enough provenance information to help govs know where the data come from and state the licence of theirs. This could be addressed by  picking up the related ontology terms. Depending how we concretely use the keyword "pragmatic", even a limited subset of the provenance ontology(ies) may be necessary.

"Pragmatic" is a great term and makes us feel good, but we need to pin
it down ;)

There is some overlap between these points (e.g. where the data came
from, how it was converted, who converted it, the license under which
it was published, etc) in the PROV example at
<http://www.w3.org/2011/prov/wiki/ProvenanceExample>

What we need to do in the Best Practices discussion is "contextualize"
such examples...

> That said, nothing prevents the "Versioning" part to also use the result from the provenance WG if they produce some recommendations that are useful to track versioning of data sets. If so, provenance would be added to both Vocabularies and Versioning ;-)

I think I agree; I think that part of "Versioning" is a view or filter
on "Provenance." Something to pay particular attention to in
"Versioning" is naming

Note: The RPI LOGD portal is currently down for maintenance; as soon
as it is back up I can provide some examples of how we generate
provenance metadata for the datasets we convert and publish.


-- 
John S. Erickson, Ph.D.
Director, Web Science Operations
Tetherless World Constellation (RPI)
<http://tw.rpi.edu> <olyerickson@gmail.com>
Twitter & Skype: olyerickson

Received on Tuesday, 27 September 2011 13:50:14 UTC