Print

Print
Hi Joe,

As I see it, the introduction of an Item class in the BIBFRAME 2.0 proposals [1] goes a long way toward meeting the library use cases we typically think of WEMI serving. With the BIBFRAME Work/Instance/Item structures in place we can begin to say that one BIBFRAME Work is the expressionOf another Work, which has a print or electronic bf:Instance, and that Instance has a particular copy (a bf:Item) that is bf:heldBy a Library.

In the above scenario, bf:Works are a combination of FRBR Works and Expressions (content), bf:Instances align closely with Manifestations (carriers of content), and bf:Items to FRBR Items (single Copies of Carriers). I actually think this is a feature of BIBFRAME; it aligns pretty well with an ontology design pattern already vetted by the semantic web community [2]. It also is close to the schema:CreativeWork/schema:ProductModel/schema:IndividualProduct “trinity”. Also, the distinctions between FRBR Works and Expressions are notoriously difficult to maintain; there are whole books on this [3]. By focusing more generally on how some particular content is the expression of other content, we end up more user focused. The *conceptual* language-neutral FRBR Work might then be represented by the graph of related bf:Works. 

Related, the RDA Unconstrained Properties [4][5] are an acknowledgement from the RDA developers that our models will need to be less stringent if we want to align our data with other data on the web. BIBFRAME’s looser definitions are similar means to the same ends. So having said that… I could imagine a community of practice around BIBFRAME that uses some combination of RDA’s Unconstrained Properties and other existing domain-specific ontologies [6] [7] [8]. 

There may be whole parts of RDA that we no longer need to employ (fingers crossed that AAPs go away). RIMFF shows us that URIs and the concatenation of atomic data points can free us from having to create unique lexical labels for everything.

Respectfully,
Steven

————
Steven Folsom
Metadata Technologies Program Manager
Harvard Library
http://orcid.org/0000-0003-3427-5769 


[1] https://www.loc.gov/bibframe/docs/pdf/bf2-draftspecitems-10-29-2015.pdf
[2] http://ontologydesignpatterns.org/wiki/Submissions:Information_realization 
[3] http://kcoyle.net/beforeAndAfter/index.html
[4] http://www.rdaregistry.info/Elements/u
[5] https://link-springer-com.ezp-prod1.hul.harvard.edu/chapter/10.1007/978-3-319-24129-6_13 
[6] http://musicontology.com/
[7] http://motools.sourceforge.net/event/event.html
[8] http://www.geonames.org/ontology/documentation.html#Map


From: Program for Cooperative Cataloging <[log in to unmask]> on behalf of Kate Harcourt <[log in to unmask]>
Reply-To: Program for Cooperative Cataloging <[log in to unmask]>
Date: Monday, March 21, 2016 at 4:38 PM
To: "[log in to unmask]" <[log in to unmask]>
Subject: Re: [PCCLIST] RDA and BIBFRAME

Thank you Joe for your recent post to the PCClist.   We are very interested in making more use of PCClist to address community concerns and questions about BIBFRAME and the Strategic Directions and are happy for this opportunity to respond.

The PCC Steering Committee discussed the issues you raised and agree that there is tension that needs to be considered but believe it is still too early in the process for any resolution to occur.  Not enough work has been done for us to imagine all the ways in which cataloging will change. The PCC and LC  both are actively engaged in monitoring the LC BIBFRAME pilot.  Data from the pilot was produced using RDA and it will be carefully reviewed and analyzed for sufficiency by LC, PoCo and other stakeholders.  Several PoCo members are also part of the anticipated LD4P/L projects and it is expected that much will be learned once work begins.  The combination of production work and research should answer many questions.  PCC is also about to announce several new committees and task groups which will help increase community understanding and perhaps more importantly bring more catalogers into the conversation.  The May OpCo meeting will have agenda topics devoted to the issues around RDA and the transition to LOD and has even added an extra half day for a "Moving Away from MARC-athon" to raise awareness.  We know that many catalogers are eager to get started and the amount of time it takes to lay the groundwork and develop consensus is frustrating.  PCC's role at this point is to learn, engage in discussion and experiment so that in the future we will be able to exercise our traditional role in standards development and training in best practices.

Please continue to raise questions as this will help move the process forward.

Kate




On 3/15/2016 10:42 AM, Joseph Kiegel wrote:
[log in to unmask]" type="cite">

The relationship between RDA and BIBFRAME needs a closer look by the PCC.

 

PCC adopted RDA for authorities in 2013 and for bibliographic data in 2015.  PCC has also promulgated many RDA guidelines, as well as the BIBCO RDA and CONSER RDA standard records.

 

BIBFRAME is a new way of representing and exchanging bibliographic data.  One of its principles is that it “generally aims to be independent of any particular set of cataloging rules” (http://www.loc.gov/bibframe/faqs/#q04).  In other words, BIBFRAME includes many but not all RDA elements, and by design will not include them all.  As an example, consider support for relationship designators in RDA and BIBFRAME:  a mapping of Work designators from RDA to BIBFRAME (http://www.lib.washington.edu/msd/pubcat/ld/resource-work) shows that a great deal of specificity is lost in BIBFRAME.

 

PCC support for both RDA and BIBFRAME creates a tension that must be resolved.  If, on one hand, RDA is the PCC content standard, why should RDA data be encoded in a scheme that irretrievably loses many intellectual distinctions made by the cataloging code?  If, on the other hand, BIBFRAME is the PCC encoding standard, why should members pay catalogers to make distinctions that are immediately discarded?

 

To my mind, our content standard is of primary importance.  Its rules embody the intellectual added value that catalogers bring as they make sense of a complex bibliographic world.  Once a content standard is chosen, then a linked-data encoding scheme should be selected that is capable of handling the full range of elements in the content model.  It is the content model that is chosen first, not the encoding scheme.

 

In my view, PCC needs to make a commitment to a content standard and to a linked-data encoding scheme that support completely loss-less data transfer.  Note that the new environment does not require the use of a single encoding scheme.  Different schemes support different use cases and BIBFRAME has its place.  But for the use case of library-to-library exchange of RDA cataloging, PCC needs a loss-less encoding scheme.

 



-- 
Kate Harcourt
Director, Original and Special Materials Cataloging
102 Butler Library
Columbia University
New York, NY  10027
http://orcid.org/0000-0003-0370-1796
email: [log in to unmask]
phone: 212.854.2714