[open-bibliography] New BNB sample data available

[Antoine Isaac]

Hi Corine,

[ I'm ccing the library linked data list: this begins to smell like real *linked* data that has tricky LD issues ;-)  ]

It's really great if BNB explicitly includes links to LD sources like LCSH!  It deserves feedback indeed...

The way you make the connection puzzles me a bit, though. Taking the following example:

   <rdf:Description>
     <dcterms:title>London fragments : a literary expedition</dcterms:title>
     <dcterms:alternative>Londoner Fragmente. English</dcterms:alternative>
[...]
     <dcterms:subject>
       <rdf:Description>
         <skos:inScheme rdf:resource="http://id.loc.gov/authorities#conceptScheme" />
         <skos:prefLabel>Görner, Rüdiger--Travel--England--London.</skos:prefLabel>
         <rdf:type rdf:resource="http://www.w3.org/2004/02/skos/core#Concept" />
       </rdf:Description>
     </dcterms:subject>
     <dcterms:subject>
       <rdf:Description rdf:about="http://id.loc.gov/authorities/sh2008107012#concept">
         <skos:inScheme rdf:resource="http://id.loc.gov/authorities#conceptScheme" />
         <skos:prefLabel>Literary landmarks--England--London.</skos:prefLabel>
         <rdf:type rdf:resource="http://www.w3.org/2004/02/skos/core#Concept" />
       </rdf:Description>
     </dcterms:subject>


I have tried in the past to exploit the record data and make connection between with existing concepts from other sources, and of course there are many problems (because of pre-coordination, or just because of typos or label changes in the considered KOS).
So I understand why you define "on-the-fly" (and "in-the-data") the concepts that you can't find in the LCSH linked data. And I think this is a reasonable solution.

What puzzles me is the second subject, for which you could recognize an existing concept from id.loc.gov. Here you duplicate the data, by copying some info you have. I'd have expected
<dcterms:subject rdf:resource="http://id.loc.gov/authorities/sh2008107012#concept">
alone.

Here you are in effect duplicating data that is served on id.loc.gov. Is it to give some minimum data that you want to provide applications with, so that they would not have to go and fetch data from id.loc.gov, a sort of "data cache"?

Also, is it generated only from the original book record at BL only? If yes, there are two risks:

- that your data is less complete than the one of other services [1], but still it lets your data consumer think that this it is complete. In which case, these consumers could decide to have their services not follow their nose to the more complete data, which could be harmful to everyone.

- that your data conflicts with the reference one. This is in fact the case, as your prefLabel ends with a period ("Literary landmarks--England--London.") while the ones at id.loc.gov do not [2]. Ok, here I am quite nitpicking, but there could be cases where the mismatch is bigger.

Best,

Antoine

[1] e.g., if you don't have skos:broader that id.loc.gov has for LCSH concepts, or the xml:lang tag with "en" for your skos:prefLabel
[2] http://id.loc.gov/authorities/sh2008107012.rdf

> ----------------------------------------------------------------------
>
> Message: 1
> Date: Tue, 1 Feb 2011 16:19:41 -0000
> From: "Deliot, Corine"<Corine.Deliot at bl.uk>
> Subject: [open-bibliography] New BNB sample data available
> To:<open-bibliography at lists.okfn.org>
> Message-ID:
> 	<19BB7EB59CBC594D8DD9F3ACA1EB24A201538E6A at w2k3-bspex2.ad.bl.uk>
> Content-Type: text/plain; charset="us-ascii"
>
> This is to let you know that there are two new sample data files
> available from our website
>
> http://www.bl.uk/bibliographic/datasamples.html
>
>
>
> The first file is an updated version of the BNB in RDF/XML. The
> substantive change is that the conversion now carries over the MARC
> country code for the place of publication.
>
>
>
> The second file is based on the same conversion but includes links to
> linked data sources: LCSH in SKOS, MARC country and language codes,
> Dewey info, Lexvo, GeoNames and the RDF Book Mashup.
>
>
>
> Feedback welcome.
>
>
>
> Cheers
>
>
>
> Corine
>
>
>
>
>
> Corine Deliot
>
> Metadata Standards Analyst
>
> British Library
>
> Boston Spa, Wetherby
>
> West Yorkshire LS 23 7BQ
>