[Phantom Landscapes]

#Linked data, BibFrame and #libraries

Posted on May 7, 2017 Updated on May 7, 2017

I spent thursday and friday (04-05 may 2017) at the Library of Congress hearing about how the PCC committees are working with linked-data for libraries.

PCC (Program for Cooperative Cataloging) is a Library of Congress hosted standards group made of mostly academic libraries in the areas of metadata, marc21 and cataloging in general. The general sub areas of PCC are BIBCO (for records standards), NACO (for name authority data), CONSER (for serials handling), and SACO (for subject creation & management). The library at which I am Head of Cataloging/Metadata is a PCC member – so this is important for my professional development etc.

I am not really here to extol the virtues of standards in marc-based cataloging – as that could take a while. I am a fan. I am actually here to talk a little about how PCC libraries are examining linked-data for cataloging and all the necessary tools required to work with data in this form.

One surprising fact that I had not considered during my ongoing learning about linked data is the lack of totality for bibframe to contain the necessary modules and preset connections to all levels of vocabularies and thesauri related to material forms, name authority files and the like. Within a Library of Congress context, their catalogers can link to the files they have already defined as most of the necessary authority work and terms are their standard files anyway. Not so with everyone else.

Differences of authority data** depend on the nature of the community being served by the library doing the work. What I mean is that there are lots of linked data files relative to names – each with their own rules and biases – from VIAF to ISNI. VIAF brings together several standard name authority services maintained globally while ISNI is a subset of name authority linked services. Whether one or the other is used depends on the context.

OCLC, which hosts VIAF, uses its traditions of global connections over the internet to aggregate name authority data into one place. One challenge to this approach is that it is not clear that every library is interested in just any name authority file. So if a library wants to use one authority file over another, they will need to link to that one specifically. The open-ness of linked-data creates new challenges in metadata sources that do not exist to such a degree in marc.

This challenge pertains to genre/form headings as much as to authority information. There are a range of terms a cataloger or metadata specialist can attach to descriptions for resources and objects in collections, but not all these come from Library of Congress. They do have their extensive genre/form vocabulary ($2 lcgft), but there is also Getty’s Art and Architecture Thesaurus. The correct list used will just depend on what the local library needs for its users and how it gains access to these files over the internet.

The current web environment opens the doors to entirely new connections of computers to computers – exposing files and data sets to each other across the network. Way different approach to the current model – which is built on a strict & contained record. The new challenges for libraries and linked-data (in bibframe) reveal that the configurations possible are myriad – which means that there are a number of real infrastructure changes required to use these metadata sources for locally maintained catalogs.

This observation was just one of the many learned notions at the PCC meetings.

Thank you for reading.

** here, authority data refers to a standard file form for a name of a person, corporate entity or family name that is used so that all materials or results of queries for searched names can return all (at least theoretically) such materials and resources that might be relevant – either in a local library search, or even more important, in an open-web environment.