Print

Print


As the 1st thread notes, dcterms:identifier was intentionally designed to have a range of rdfs:Literal rather than rdfs:Resource. This is a common situation in local datasets where codes or numbers are used as identifiers rather than URIs. Within the context of that dataset, there can be a presumption of uniqueness. This assumption generally fails, though, when the data is aggregated and the context is typically ignore (e.g. N-Triple downloads).

 

In contrast, owl:sameAs performs the same function, but doesn’t lose context in aggregations.

 

Jeff

 

From: Bibliographic Framework Transition Initiative Forum <[log in to unmask]> on behalf of "Deliot, Corine" <[log in to unmask]>
Reply-To: Bibliographic Framework Transition Initiative Forum <[log in to unmask]>
Date: Friday, March 1, 2019 at 6:27 AM
To: "[log in to unmask]" <[log in to unmask]>
Subject: Re: [BIBFRAME] When a bf:Identifier is a URI

 

Hello,

 

I would like to pick up on Jeff’s comment in the thread below “Is it BIBFRAME's intention to create a barrier from the linked data cloud by treating URIs as literals?

 

A long time ago (2010), our first open data release in RDF/XML sparked a discussion about recording URIs as strings.

 

The discussion starts here: https://lists.okfn.org/pipermail/open-bibliography/2010-November/000623.html and concludes here: https://lists.okfn.org/pipermail/open-bibliography/2010-November/000628.html (short thread; takes 3 clicks to go from one to the other).

 

The conclusion of that exchange seemed to be that it was not a good idea to encode URIs as literals, even if it was “valid RDF” and “conform[ing] to the defined semantics” apart from “some obscure cases involving reification” . Isn’t “encoding URIs as literals” what part of this discussion is about, beyond Bibframe, issn, cool URIs, URNs, etc.? And if  so, what do people on this list think?

 

Thank you.

 

Corine

 

Corine Deliot

Collection Metadata Analyst

The British Library

From: Bibliographic Framework Transition Initiative Forum <[log in to unmask]> On Behalf Of Young,Jeff (OR)
Sent: 28 February 2019 17:58
To: [log in to unmask]
Subject: Re: [BIBFRAME] When a bf:Identifier is a URI

 

My impression is that bf:identifiedBy exists because their is discomfort using owl:sameAs like so:

 

<URI for LC work> owl:sameAs <URI for ISSN work> .

 

The idea seems to be that adding a level of indirection is a way to hedge bets:

 

<URI for LC work> bf:identifiedBy "http://issn.org/..."^^xsd:URI .

or

<URI for LC work> bf:identifiedBy [a bf:URI; rdf:value "http://issn.org/..."]

or

something similar.

 

There are other ways to hedge the bet, though, that don't involve encoding the URI as a literal or treating it as an rdf:value. One way would be to treat bf:identifiedBy as equivalent to umbel:isLike:

 

http://techwiki.umbel.org/index.php/UMBEL_Vocabulary

 

This hedge would only be necessary, though, in cases where the data publisher was unsure of the sameness.

 

Another solution is how Wikidata does it, with different properties for each external dataset. This would allow for a default assumption of sameness, while exceptions to be declared on a reified node when unsure (e.g. when the sameness is disputed or mistaken associations have been deprecated.)

 

I can find examples in Wikidata if folks want to look closer at that solution.

 

Jeff

 

 


From: Bibliographic Framework Transition Initiative Forum <[log in to unmask]> on behalf of Ray Denenberg <[log in to unmask]>
Sent: Thursday, February 28, 2019 12:40 PM
To: [log in to unmask]
Subject: Re: [BIBFRAME] When a bf:Identifier is a URI

 

Jeff, I think we could address this issue better if we had a real live use case, which I still don’t think we have.  I believe we are in agreement that you don’t want to supply the issn for a resource via the URI for that issn using bf:identifiedBy, but rather by the issn itself (using bf:identifiedBy)..  Would you agree?

 

I think we have also established that if the URI resolves to a description of a Work, and it happens to be a URI for an ISSN, you CAN say:

 

          <URI for issn>  a bf:Work

 

that is, you DO use the URI form, not a quoted string.  But you don’t want to express the fact that a given URI resolves to a description of a work using bf:identifiedBy.   Would you agree?

 

So then, what is a use case for supplying a URI via bf:identifiedBy?

 

Ray




On Feb 28, 2019, at 12:17 PM, Young,Jeff (OR) <[log in to unmask]> wrote:

 

Is it BIBFRAME's intention to create a barrier from the linked data cloud by treating URIs as literals?

 

Jeff


From: Bibliographic Framework Transition Initiative Forum <[log in to unmask]> on behalf of Svensson, Lars <[log in to unmask]>
Sent: Thursday, February 28, 2019 7:43 AM
To: [log in to unmask]
Subject: Re: [BIBFRAME] When a bf:Identifier is a URI

 

Hi Ray,

On Friday, February 22, 2019 8:42 PM, Denenberg, Ray wrote:

> I’m still wondering why you’d want to do this:
> 
> some:resource bf:identifiedBy [
>          a some:type ;
>          rdf:value “ http://issn.org/resource/ISSN/1084-5305” ]
> 
> (quoted or not)
> 
> Vs.
> 
> some:resource bf:identifiedBy [
>          a bf:Issn ;
>          rdf:value “1084-5305” ]

You're right, Ray: I don't want to do the first when it comes to ISSNs. If, however, the canonical identifier for a resource happens to be a URI, I'd want to treat that URI as any other identifier and put it in quotes.

Best,

Lars

> 
> 
> From: Bibliographic Framework Transition Initiative Forum
> <mailto:[log in to unmask]> On Behalf Of Svensson, Lars
> Sent: Thursday, February 21, 2019 11:26 AM
> To: mailto:[log in to unmask]
> Subject: Re: [BIBFRAME] When a bf:Identifier is a URI
> 
> Dan,
> 
> Re your question if „the requirement to quote the ISSN URI as a string when used
> as the bf:Issn object of bf:identifiedBy hold true?” :
> 
> My take would be that an instance of the class bf:Issn cannot be a quoted string
> since that would lead to a literal in the subject position:
> 
> some:resource bf:identifiedBy “http://issn.org/resource/ISSN/1234-5678#ISSN” .
> 
> which implies
> 
> “http://issn.org/resource/ISSN/1234-5678#ISSN” a bf:Issn .
> 
> since we have
> 
> bf:Issn rdfs:subClassOf bf:Identifier.
> 
> If we on the other hand have
> 
> some:resource bf:identifiedBy [
>          a bf:Issn ;
>          rdf:value “http://issn.org/resource/ISSN/1234-5678#ISSN” .
> ] .
> 
> we should be fine.
> 
> Or did I misunderstand your proposal?
> 
> Best,
> 
> Lars
> 
> *** Lesen. Hören. Wissen. Deutsche Nationalbibliothek ***
> --
> Dr. Lars G. Svensson
> Deutsche Nationalbibliothek
> Informationsinfrastruktur
> Adickesallee 1
> 60322 Frankfurt am Main
> Telefon: +49 69 1525-1752
> Telefax: +49 69 1525-1799
> mailto:[log in to unmask]
> http://www.dnb.de
> 
> From: Bibliographic Framework Transition Initiative Forum
> [mailto:[log in to unmask]] On Behalf Of Dan Scott
> Sent: Wednesday, February 20, 2019 11:45 PM
> To: mailto:[log in to unmask]
> Subject: Re: [BIBFRAME] When a bf:Identifier is a URI
> 
> Ray et al:
> 
> Thank you to both the BIBFRAME mailing list and the LC BIBFRAME team for taking
> this seriously.
> 
> > Still, bibframe does not preclude declaring that a particular URI is an identifier for
> a particular resource (although we would still welcome a use case).
> 
> ISSNs have had dereferenceable URIs for almost two years now
> (https://www.issn.org/wp-content/uploads/2017/12/ISSN-
> LinkedDataApplicationProfile-v0_9.pdf). Does the requirement to quote the ISSN
> URI as a string when used as the bf:Issn object of bf:identifiedBy hold true?
> 
> Thanks,
> Dan
> 
> P.S. In a side channel, Sarven Capadisli pointed me at the 2017 document
> https://www.w3.org/DesignIssues/QuotingURIs.html which describes a rationale for
> quoting URIs in certain situations, where owl:sameAs may result in two different
> URIs no longer being distinguished properly. I found this helpful, although I don't
> think it comes into play here, as bf:identifiedBy does not make any owl:sameAs
> assertions that would trigger the detrimental outcomes outlined in the design note.
> 
> On Tue, 19 Feb 2019 at 15:58, Denenberg, Ray <mailto:[log in to unmask]> wrote:
> The LC bibframe team discussed this at length this morning.
> 
> There were only about four messages at the time and now there are now many
> more, too many to keep up with it all.  Among those early messages there seemed to
> be only this use case:
> 
> <http://example.org/2335409#Work> a bf:Text, bf:Work ;
>    bf:identifiedBy <http://worldcat.org/entity/work/id/638612>.
> 
> Which seems to say, there is a work, http://example.org/2335409#Work, at
> institution A, and this, http://worldcat.org/entity/work/id/638612, at institution B,
> is the same work.
> 
> (Nevermind what constitutes sameness -  whatever institution A considers it to
> be.  Could be matched based on author and title for example.)
> 
> We don’t think this is an appropriate use of
> bf:identifiedBy.     http://worldcat.org/entity/work/id/638612    doesn't identify
> http://example.org/2335409#Work. Rather, the assertion is that they are the same
> resource.
> 
> We recognize that there is a need to make such an assertion, that bibframe
> currently does not provide one, and we will investigate this further.
> 
> Still, bibframe does not preclude declaring that a particular URI is an identifier for a
> particular resource (although we would still welcome a use case).   As has been
> suggested, it should be quoted (a string) rather than a resource.  bf:identifiedBy is
> intended for string identifiers, not for identifiers that are themselves linked data
> resources.
> 
> Example:
> 
> <http://example.org/workA>
>        bf:identifiedBy [
>                    a   xs:anyURI ;
>                   rdf:value     “http://example.org/identifierA” ]
> 
> Ray
> 
> From: Bibliographic Framework Transition Initiative Forum
> <mailto:[log in to unmask]> On Behalf Of Dan Scott
> Sent: Friday, February 15, 2019 7:45 PM
> To: mailto:[log in to unmask]
> Subject: [BIBFRAME] When a bf:Identifier is a URI
> 
> While the description for bf:Identifier
> (http://id.loc.gov/ontologies/bibframe.html#c_Identifier) is "Token or name that is
> associated with a resource, such as a URI or an ISBN", its subclasses include bf:Isbn
> but not bf:Uri[1].
> 
> For now I've interpreted this as "just use bf:Identifier if you have a URI," but if that
> interpretation is correct, it might make sense to clarify the description.
> 
> Example:
> 
> @prefix bf: <http://id.loc.gov/ontologies/bibframe/> .
> <http://example.org/2335409#Work> a bf:Text, bf:Work ;
>    bf:identifiedBy <http://worldcat.org/entity/work/id/638612>.
> 
> <http://worldcat.org/entity/work/id/638612> a bf:Identifier .

 


 
******************************************************************************************************************

Experience the British Library online at www.bl.uk

The British Library’s latest Annual Report and Accounts : www.bl.uk/aboutus/annrep/index.html

Help the British Library conserve the world's knowledge. Adopt a Book. www.bl.uk/adoptabook

The Library's St Pancras site is WiFi - enabled

*****************************************************************************************************************

The information contained in this e-mail is confidential and may be legally privileged. It is intended for the addressee(s) only. If you are not the intended recipient, please delete this e-mail and notify the [log in to unmask] : The contents of this e-mail must not be disclosed or copied without the sender's consent.

The statements and opinions expressed in this message are those of the author and do not necessarily reflect those of the British Library. The British Library does not take any responsibility for the views of the author.

*****************************************************************************************************************

Think before you print