[ckan-discuss] Confusion regarding the use of provenance vocabulary for the CKAN metadata

Fri Apr 1 13:48:06 BST 2011

On 01/04/11 13:44, William Waites wrote:
> * [2011-04-01 13:04:26 +0100] Monika Solanki<monika.solanki at gmail.com>  écrit:
>
> ] From what I read in the Open Provenance Model Vocabulary Specification at
> ]
> ] http://open-biomed.sourceforge.net/opmv/ns.html,
> ]
> ] it says  "... For the moment, we would recommend users to use the
> ] Provenance Vocabulary for describing provenance of the Web of Data. ".
> ] This Provenance vocabulary is the one available at,
>
> Depending on how one understands the term "Web of Data", CKAN includes
> datasets "beyond" it. It certainly includes non-semweb data. It may
> include data that is not directly accessible (i.e. behind a "register
> to download" page or the like). At the very least some of the more
> granular terms, indended for describing the provenance of individual
> data items are not applicable to most of the datasets on CKAN, even
> the semweb ones (where what exactly a data item is might not be clearly
> defined).
>
> OPMV is also simpler than PRV for the basic case, and it is not
> incompatible with PRV, they can be used together if it is warranted.
>
> Might it make sense to revisit this choice, given that it was made well
> before the provenance IG finished its work and well before the current
> version of either of those vocabularies? Quite possibly. Is it
> inappropriate to use OPMV in this way? I don't think so.
Me neither! I was only trying to understand the rationale given the 
recommendation.

Thanks!

M
> Cheers,
> -w
>