Comments from the Netherlands on the Linked Data adoption of PBcore

[JohanOomen]

Good afternoon,

We have studied the PBCore documentation from the perspective of the Netherlands Institute for Sound and Vision. Most of the modelling decisions seem reasonable to us and indeed, most of our own metadata could be mapped to this format. I have some remaining questions and remarks:

This version of PBCore opens the door to Linked Data adoption, which is great. At S&V we are publishing data and vocabularies using Linked Data and therefore appreciate this effort. In the pdcore definition "ref" fields are used for this and they are intended for referencing internal or external resources through URIs. As they would appear inside an XML document, this is not really RDF or Linked Data yet, but a good step and it would make conversion to actual Linked Data easier. I think publishing at least the schema as 5-star linked data would pave the way for inclusion of PBCore in the semantic web (DC is one of the most used vocabularies in the Web of Data as of now).

• Are there any plans for an RDFS version of the schema (DC is available in this format).
• The properties and classes used now do have URIs which are dereferenceable. However, they only respond with HTML.
  • Are the implied URIs indeed the ones we want: http://www.pbcore.org/PBCore/pbcoreID
  • Will they (in the future) respond with RDF back? (through content negotiation)

One small modeling question I had was concerning creatorRole which would include "the interviewee from a video history program"

• Are people that appear in the video always considered "creators"?
  • For example, a dancer, choreographer in a Dance Heritage Collection video?

Finally, a minor issue: at the time of checking, the http://www.pbcore.org/v2/attributes/affiliation link is not responding.

Best regards, --Johan Oomen @johanoomen and Victor de Boer @victordeboer

[kvanmalssen]

Hi Johan and Victor,

Thanks for your comments. We are currently working on revising the data model, elements and XML schema to meet community needs. After that, we will be turning our attention to creating an RDF ontology for PBCore. I expect that we will be getting to that in early 2015. We agree, it is important to move in this direction.

Regarding your question about creators, PBCore has a "roles" element that is used with creators (and another with contributor). So in order to describe a choreographer, you would use:

<pbcoreCreator>
   <creator>Susan Marshall</creator>
   <creatorRole>choreographer</creatorRole>
</pbcoreCreator>

Does this meet the use case you have in mind?

Thanks for the comment about the affiliation link. I think this is the same issue that was raised in https://github.com/WGBH/PBCore2.0/issues/81

[AllisonAnn]

Hi - Just to add to Kara's example regarding the creator/role question

• pbcore is flexible. Depending on how you choose to catalog/rank an individual's role related to the creation of an asset, you might choose to record the individual as a contributor :

Susan Marshall choreographer

I think that the Role controlled vocab might be similar for both creator and contributor. I also think pbcore should just maintain a single Role list, and let the cataloger decide if it is creator or contributor in nature. If I haven't already suggested that - I am now.

Allison

On Fri, Sep 12, 2014 at 9:09 AM, kvanmalssen notifications@github.com wrote:

Hi Johan and Victor,

Thanks for your comments. We are currently working on revising the data model, elements and XML schema to meet community needs. After that, we will be turning our attention to creating an RDF ontology for PBCore. I expect that we will be getting to that in early 2015. We agree, it is important to move in this direction.

Regarding your question about creators, PBCore has a "roles" element that is used with creators (and another with contributor). So in order to describe a choreographer, you would use:

Susan Marshall choreographer

Does this meet the use case you have in mind?

Thanks for the comment about the affiliation link. I think this is the same issue that was raised in #81 https://github.com/WGBH/PBCore2.0/issues/81

— Reply to this email directly or view it on GitHub https://github.com/WGBH/PBCore2.0/issues/87#issuecomment-55408147.

[biktorrr]

Hi and thanks for the replies,

I guess my question/concern was more on the intended scope of the creator relation. A choreographer is probably a bit of an ambiguous case as he/she could be considered the creator or contributor of the thing depicted in a video or the work itself. But would let's say a painter depicted in a documentary video would be a subject rather than a creator. I assume the intended scope of creator / contributor is that of the video, not something (like a work) depicted? I m probably nitpicking here, as in practice it will probably be up to the discretion of the institutions.

best --victor

[AllisonAnn]

Hi Victor -

In regards to the pbcore Subject fields, it would be useful to have a type vocab that would include the term "depicted" for such cases. Personally, I would link the person depicted in the work as a subject, and not a creator or contributor...

Allison

On Mon, Sep 15, 2014 at 2:07 AM, Victor de Boer notifications@github.com wrote:

Hi and thanks for the replies,

I guess my question/concern was more on the intended scope of the creator relation. A choreographer is probably a bit of an ambiguous case as he/she could be considered the creator or contributor of the thing depicted in a video or the work itself. But would let's say a painter depicted in a documentary video would be a subject rather than a creator. I assume the intended scope of creator / contributor is that of the video, not something (like a work) depicted? I m probably nitpicking here, as in practice it will probably be up to the discretion of the institutions.

best --victor

— Reply to this email directly or view it on GitHub https://github.com/WGBH/PBCore2.0/issues/87#issuecomment-55558321.