[DCRM-L] FW: Frequently Asked Questions ? Library of Congress Linked Open Data Hosting Solution

Mon Jul 30 10:29:01 MDT 2018

Thanks Amy, I'm glad that this is open to discussion, and sorry if I have already missed any threads about subdivision.  Speaking for myself, I'm not wed to subdivisions unless they relate to vocabularies not directly related to the specific publication/manuscript.  But I fear that this would cover most of the RBMS vocabularies.  I would want to be able to subdivide information about a binding when I know it, or a manuscript cast list, or annotations, etc.  These are difficult enough to retrieve in many library databases even WITH the subdivisions, so I hate to lose those options.

That said, I do recognize the importance of creating LOD for our vocabularies.  Will the LDS allow a way to link to part of a string, as one can in Connexion?  Perhaps that might answer the problem,
Good luck,
andrea

-----Original Message-----
From: DCRM-L [mailto:dcrm-l-bounces at lib.byu.edu] On Behalf Of dcrm-l-request at lib.byu.edu
Sent: Thursday, July 26, 2018 10:10 PM
To: dcrm-l at lib.byu.edu
Subject: DCRM-L Digest, Vol 149, Issue 12

Today's Topics:

   1. Re: FW: Frequently Asked Questions ? Library of Congress
      Linked Open Data Hosting Solution (Amy Brown)

----------------------------------------------------------------------

Message: 1
Date: Thu, 26 Jul 2018 22:09:16 -0400
From: Amy Brown <amy.brown.3 at bc.edu>
To: "DCRM Users' Group" <dcrm-l at lib.byu.edu>
Subject: Re: [DCRM-L] FW: Frequently Asked Questions ? Library of
	Congress Linked Open Data Hosting Solution
Message-ID:
	<CAFX2NKwudVezSYXhL1Er3wN+n7_5GSWzyZKwcbLADvUt1nKPcA at mail.gmail.com>
Content-Type: text/plain; charset="utf-8"

Hi Andrea:

This is an excellent question, and one that the CVEG needs to give continued thought to as we proceed.

The short answer is maybe. There's nothing about LOD *per se* that prohibits a subdivided vocabulary. Where it gets complicated for the RBMS CV is that, as of now, the Vocabularies consider the subdivisions solely as part of the application instructions. That is, we don't actually maintain them as separate terms, which could be problematic.

So, we need to investigate further before we can say for certain how/if subdivision will be feasible going forward. In the interim, I'd be interested in hearing more from the community about the level of need.

Best,

Amy Brown
CV co-editor

On Thu, Jul 26, 2018 at 6:04 PM, Cawelti, Andrea <cawelti at fas.harvard.edu>
wrote:

> This was a nice FAQ about the future of the RBMS controlled vocab 
> project, thanks to all for sharing it.  A question, will we still be 
> able to add $z and $y to our terms, once they go 2.0?  I?m 
> particularly concerned about the provenance-related terms.  I?m still 
> struggling with the nuances of LOD, so it?s entirely possible that I?m not understanding something here.
>
> Enquiring minds want to know,
>
> andrea
>
> --
>
> (Ms.) Andrea Cawelti
> Ward Music Cataloger
> Houghton Library
> Harvard University
> Cambridge, MA  02138
>
> Phone: (617) 998-5259
> FAX: (617) 495-1376
> E-mail: cawelti at fas.harvard.edu
>
>
>
>
>
>
>
> *Feed:* RBMS Controlled Vocabularies Community Discussion *Posted on:* 
> Thursday, July 26, 2018 1:58 PM
> *Author:* amyfbrown
> *Subject:* Frequently Asked Questions ? Library of Congress Linked 
> Open Data Hosting Solution
>
>
> The RBMS Controlled Vocabularies Editorial Group (CVEG) seeks feedback 
> on a partnership with the Library of Congress Linked Data Service 
> (LDS) to host the RBMS Controlled Vocabularies (RBMS CV). The CVEG has 
> been investigating options for several years for making the RBMS CV 
> available as Linked Open Data (LOD). Because of the technological 
> infrastructure requirements, we are unable to host the vocabularies as 
> LOD ourselves and have sought out partnerships to pursue this 
> important initiative. Below are some frequently asked questions about this project.
>
>
>
> *What is Linked Open Data (LOD)?*
>
>
>
> Releasing an authority file as Linked Open Data has a few basic 
> requirements. Each thing, which in our case are individual terms, must:
>
>    - Have a stable, unique, dereferenceable* URI
>    - Use a URI that provides useful information when queried by either
>    humans or machines.
>    - Use open data standards, such as RDF
>    <https://en.wikipedia.org/wiki/Resource_Description_Framework> or
>    SPARQL <https://en.wikipedia.org/wiki/SPARQL>
>    - Refer to other things when published
>
> Linked Open Data makes traditional authority control much easier by 
> relying on stable identifiers that point to a thing, rather than a string.
> For instance, in linked open data, the URI actually refers to a 
> conceptual understanding of ?vellum bindings,? as opposed to simply 
> listing the characters that compose the term ?vellum bindings.? Using 
> these URIs, developers can change preferred terms immediately, without 
> having to manually change the character strings displayed for users.
>
> Relying on encoded relationships, LOD allows for searching, 
> collocation and discovery across vocabularies, systems and encoding 
> standards. In systems that take full advantage of LOD, information or 
> examples of certain concepts are easily findable and understandable by 
> machines, which translates to easy findability of a resource via major 
> search engines by people.
> *Note: Dereferenceable URIs return human-readable information when 
> queried through an http service. In other words, a dereferenceable URI 
> is one that you can point your browser to, and read the resulting web page.
>
>
>
> *Why should we release the RBMS CV as Linked Open Data?*
>
>
>
> Perhaps the most pressing reason is that releasing the RBMS CV as 
> linked open data allows the vocabulary to be used more widely than by 
> the cataloging community and outside of the MARC cataloging 
> environment. While this will be essential for BIBFRAME, it also 
> enables easier use of the RBMS CV for other digital scholarship projects that exist outside MARC.
>
>
>
> As mentioned above, this project will also facilitate a higher level 
> of authority control than was previously feasible with the RBMS CV. 
> The RBMS CV will be available for download in a number of easily 
> manipulatable formats (RDF/XML, N-triples, and JSON) that can be 
> utilized by most major authority control vendors.
>
>
>
> Currently, Library of Congress vocabularies (including LCSH, LCGFT, 
> and LCDGT), as well as AAT, are available as LOD. Releasing the RBMS 
> CV data as LOD will allow for related concepts in different 
> vocabularies to be linked, thus improving overall discovery of library resources.
>
>
>
> *How will I navigate the RBMS CV on the Library of Congress site?*
>
>
>
> Navigation of the vocabulary will be similar to other vocabularies 
> available on the id.loc.gov site. To experiment, simply try searching 
> a term here <https://id.loc.gov/>. You can limit your search to a 
> specific vocabulary by selecting it in the list below the search box. 
> After implementation, you will be able to download the entire 
> vocabulary from the ?available datasets? section.
>
>
>
> *Who will maintain the content of the RBMS CV?*
>
>
>
> The Controlled Vocabularies Editorial Group, a committee of the 
> Association of College and Research Libraries, a division of the 
> American Libraries Association, will continue to maintain editorial 
> control of the RBMS CV. We will provide regular data dumps, probably 
> quarterly, to the LC LDS, who will in turn host them on id.loc.gov.
>
>
>
> *Will the terms change?*
>
>
>
> No, the terms will not change. Concurrently, CVEG is working on a 
> project to integrate and reorganize the RBMS CV. This project will 
> result in new hierarchies and eliminate the need for parenthetical 
> qualifiers except in a few instances. The group will be providing 
> updates on our blog as this project proceeds.
>
>
>
> *Why not continue to host the RBMS CV as we do now?*
>
>
>
> The current site is available in a simple html format. While this 
> method provides a nice browsing interface, it makes even traditional 
> library authority control difficult. Dynamic use outside of the MARC 
> environment is nearly impossible.
>
>
>
> As we strive to retain relevance and longevity over time, it is 
> important for the RBMS CV to continue to adapt to changes in 
> technologies and evolving standards, particularly given the Library of 
> Congress? continued commitment to BIBFRAME.
>
>
>
> *What other options have you considered?*
>
>
>
> A lot! We looked into other partnerships, as well as hosting the 
> vocabulary as LOD ourselves on rbms.info. Unfortunately, our website 
> isn?t up to the challenge of providing both machines and humans the 
> ?useful information? required to be LOD. Technically speaking, this 
> means that the combination of our website software and server won?t be 
> able to return machine-readable data formats (such as RDF or JSON) when queried.
>
>
>
> We also investigated other partnerships, and determined that the 
> Library of Congress offers the best mix of usability, reliability and 
> longevity available.
>
>
>
> *What will happen to the current rbms.info/vocabularies 
> <http://rbms.info/vocabularies> site?*
>
>
>
> Because publication of LOD requires a single, stable URI to identify 
> each thing, or, in our case, term, we will be retiring the current 
> rbms.info site. This way, there will only be one referenceable http 
> address available for each term.
>
>
>
> As we proceed with integration, we are investigating the best methods 
> of archiving the un-integrated thesaurus for authority control and 
> record-keeping purposes.
>
>
>
> *How long-term is this solution?*
>
>
>
> Although it is impossible to predict with absolute certainty the 
> technological environment in the future, one requirement of LOD is 
> that the URIs be stable over time. For this reason, hosting by LC LDS 
> should be considered a near-permanent solution. It would be very 
> detrimental to switch publication solutions once the URIs are minted.
>
>
>
> *What is the proposed implementation date for this project?*
>
>
>
> We do not have a specific date yet. However, implementation will 
> necessarily take place after the integration and reorganization 
> project is complete.
>
>
> Have other questions? Concerns? Feedback? Please let us know! We will 
> be happy to receive comments here, or via e-mail at vocabularies at rbms.info.
>
>
> View article...
> <http://rbms.info/cv-comments/2018/07/26/frequently-asked-questions-li
> brary-of-congress-linked-open-data-hosting-solution/>
>

--
Amy F. Brown
Special Collections Technical Services, Head John J. Burns Library
140 Commonwealth Avenue
Chestnut Hill, MA  02467
617-552-8124
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserver.lib.byu.edu/pipermail/dcrm-l/attachments/20180726/ea3ce2d7/attachment.html>

End of DCRM-L Digest, Vol 149, Issue 12
***************************************