RE: changing my mind about versioning metadata

Dear Karen, dear Annette,

How to represent "metadata about metadata" is an issue that has been discussed also in the work on DCAT-AP / GeoDCAT-AP. The identified requirements didn't cover versioning, but rather issues concerning (meta)data provenance (who created / is maintaining the record, whether the record has been derived / transformed from another one) and (meta)data quality (whether a metadata record is conformant with a given metadata standard / schema / profile). All this information is specified via statements having dcat:CatalogRecord as subject (so, it is specified "inside" the metadata).

Another example is ISO 19115, which has a bunch of properties associated with metadata records - including metadata point of contact, language, char encoding, and conformity with a standard.

It is true that in many cases this could be considered administrative information (and probably metadata versioning fits just here). However, they may also address requirements concerning metadata consumption and cross-catalogue harvesting - as in the DCAT-AP / GeoDCAT-AP / ISO 19115 cases. 

I don't think we have specific use cases and requirements for metadata records (although UC30 [1] indirectly covers some aspects). However, metadata are, after all, data. So, we could consider adding just 1 use case, showing how (at least some of) the requirements identified for data, apply to metadata as well. I can help drafting it, if the WG thinks this makes sense.

BTW, it is worth mentioning that "metadata quality" is something that was also discussed in the framework of the Data Quality Vocabulary (DQV) (see mail thread starting at [2]), and that DQV includes a specific example on this:




Andrea Perego, Ph.D.
Scientific / Technical Project Officer
European Commission DG JRC
Directorate B - Growth and Innovation
Unit B6 - Digital Economy
Via E. Fermi, 2749 - TP 262
21027 Ispra VA, Italy

The views expressed are purely those of the writer and may
not in any circumstances be regarded as stating an official
position of the European Commission.

>-----Original Message-----
>From: Karen Coyle []
>Sent: Tuesday, September 26, 2017 5:19 PM
>Subject: Re: changing my mind about versioning metadata
>I think the question is whether the version information is part of the
>metadata itself or is information about the metadata that is outside of
>the metadata. (I'm assuming that DCAT will not define the content of the
>metadata - if that assumption is wrong, ignore this.) Most metadata
>records have some version information as administrative metadata, along
>with who created the metadata and other such details. Is it desirable to
>have metadata administrative properties within DCAT rather than/in
>addition to the administrative metadata of the metadata itself?
>It may help to identify the purposes of versioning for each resource.
>One purpose that comes to mind is to make it possible to retrieve the
>matching version for each resource, i.e. to obtain the the metadata for
>the dataset version that is being retrieved. This could be achieved
>through a direct link within the DCAT properties, which may be simpler
>than a coordination of versions, especially if they dataset and metadata
>don't always change in sync.
>On 9/25/17 4:37 PM, Annette Greiner wrote:
>> Thinking about this more, and looking again at DCAT, I think we do need
>> to include versioning of metadata. Since DCAT already offers
>> dct:modified as a possible attribute of a CatalogRecord, I think that it
>> would be logical to allow people to apply versioning as well.
>> -Annette
>> On 9/25/17 8:02 AM, Dataset Exchange Working Group Issue Tracker wrote:
>>> dxwg-ACTION-44: Go back to use cases to discover which resources need
>>> versioning, and create specific requirements

>>> Assigned to: Alejandra Gonzalez Beltran
>Karen Coyle

>m: 1-510-435-8234 (Signal)
>skype: kcoylenet/+1-510-984-3600

Received on Tuesday, 26 September 2017 18:19:50 UTC