The point I was trying to show -- anything can have a URI that we might want to know the provenance of, including  http://bibframe.org/resources/Ahx1405278232/1706459title7 or any other.

Thus for every resource in the model (other than bf:Identifier, which should not have identity, as previously) there should be the possibility to give bf:identifier with an instance of bf:Identifier with the URI reified in bf:identifierValue, and the provenance description in other associated properties.

To make a couple of concrete examples:

1.  We have a post-processing script that generates URIs and replaces the ones from the XQuery conversion utility.  That post-processing script is versioned, and I found a bug in it recently.  If we are going down the route of identifier provenance, then I could query to find all the URIs generated by that script.  

2. As per Simeon's example, Stanford might mint a URI for me and Cornell might mint a URI for me.  It might be interesting to know information about the assignment of those URIs... but they're not magical somehow, they're just URIs.

Currently both are supported by the model (bf:identifier's domain is bf:Resource, which has subclass of bf:Title).

Rob



On Thu, Jul 24, 2014 at 2:34 PM, Denenberg, Ray <[log in to unmask]> wrote:

Are you agreeing, then, that the provenance of URIs (of any URI scheme) is out of scope?”

 

I don’t know.  

 

Are you asking, do we need provenance for  URIs like http://bibframe.org/resources/Ahx1405278232/1706459title7    ?

 

No.

 

Ray

 

 

From: Bibliographic Framework Transition Initiative Forum [mailto:[log in to unmask]] On Behalf Of Robert Sanderson
Sent: Thursday, July 24, 2014 5:19 PM


To: [log in to unmask]
Subject: Re: [BIBFRAME] BibFrame and Linked Data: Identifiers

 

 

Hi Ray,

On Thu, Jul 24, 2014 at 2:16 PM, Denenberg, Ray <[log in to unmask]> wrote:

One can say that http://bibframe.org/resources/Ahx1405278232/1706459title7 is an identifier, and by definition it is, but
I don't believe it is an identifier in the context of the discussion we are having.  We've been talking about non-URI identifiers, that have a scheme, and can (in some/most/all ?) case be turned into URIs (http or otherwise).

 

Are you agreeing, then, that the provenance of URIs (of any URI scheme) is out of scope?

 

Rob

 

 




--
Rob Sanderson
Technology Collaboration Facilitator
Digital Library Systems and Services
Stanford, CA 94305