BIBFRAME and BIBFRAME (Lite): A Deeper Dive

During the NISO virtual conference on BIBFRAME & Real World Applications of Linked Data, Carolyn Hansen gave a clear and informative presentation on BIBFRAME, giving a general overview as well as addressing some challenges and recent developments.

She began by reviewing the original vision for BIBFRAME as well as a rough timeline of development so far. One important point to remember is that BIBFRAME is more than just "MARC 2.0" - it aims to repackage information and focus on relationships between resources rather than collating all the information about a resource with no need for context. Next Hansen moved into a description of the original BIBFRAME 1.0 model, with its four core classes of Work, Instance, Authority, and Annotation. She presented some critiques of the model, quoting Rob Sanderson in his direct comparison between Linked Data best practices and the recommendations of BIBFRAME. BIBFRAME 1.0 did not score particularly well on many of the dimensions assessed, which partly inspired the rethinking that led to BIBFRAME 2.0.

Hansen gave a very clear and thorough introduction to the new model, which now has three core classes: Work, Instance, and Item. She walked through each of these, explaining similarities and differences to the model 1.0. She also addressed the three important "key concepts" of Agent, Event, and Subject. Each of these has its own rules and properties, which she explained quite neatly. She also discussed the three methods for expressing roles in BIBFRAME 2.0: 1) role as a property, 2) role as a property with a label for the agent, and 3) role as external property (drawing on, say, id.loc.gov instead of bibframe.example.org). This gets a bit technical, but if you're interested it is well worth a listen. Hansen also went into some detail on things that BIBFRAME does not prescribe; for instance, the Library of Congress takes no stance on blank nodes. She concluded this section by recognizing that although BIBFRAME 2.0 is an improvement over the previous model, it is by no means perfect.

One way that practitioners are working to improve BIBFRAME is through BIBFRAME Lite, which began development early last year. The pared-down version in BIBRAME Lite provides an opportunity for users to practice using BIBFRAME and allows for quicker adaptation than LC is capable of. BIBFRAME Lite takes a modular approach, where participants can use a range of different profiles to record the information they need for specific material types and communities (for instance, there is a profile made for archives, as well as one specifically for rare materials).

Although BIBFRAME development continues apace and the models are still under testing and discussion, there are some general implementation challenges that are still an issue. These include communication between different stakeholders (e.g. museums, libraries, archives, large and small organizations, etc.), small and specialized data, the slowness of the development of standards, and issues with versioning - BIBFRAME Lite and many existing tools use BIBFRAME 1.0, not the newer BIBFRAME 2.0 model.

Hansen ended her presentation with a desire to see greater communication in future, as well as the creation and implementation of systems that make BIBFRAME as easy as possible for all sizes of institutions.

If you're interested in learning more about BBIFRAME and some real world applications of linked data, library staff in the Minitex region can view this and all NISO recordings via the Minitex MyLibrary portal. Just log in to your account and follow the “Archived Webinars from NISO, NASIG, ALA Editions” link. If you don’t have an account yet, visit the Minitex website, look for the “login” link in the upper right corner on the page, and follow the steps to set up an account.