The FRBR Blog

The 2006 FRBR Challenge starts on Wednesday, so come back then and see what it is. I’ll give you a hint: it involves J.R.R. Tolkien’s The Lord of the Rings. All will be revealed in two days!

Prize update: I’ll be giving out three of these S.R. Ranganathan pins too.

Some recent talk about the FRBRization of LibraryThing:

• LibraryThing at Catalogablog
• LibraryThing Reverse Engineering FRBR? at See Also; Tim Spalding from LibraryThing is in the comments and explains that the FRBRization is done by users grouping things by hand
• LibraryThing FRBRizing Through the Masses at Librarian in Black

Don’t miss the entry here a few days ago where Spalding left a lengthy comment.

Here’s a bit more background for the 2006 FRBR Challenge, in case it comes in handy. This diagram (it’s a link to a larger version, 73 KB PNG) puts three diagrams from the FRBR Report into one: it shows the Group 1, Group 2, and Group 3 entities, and the basic relationships between them.

• Group 1 entities: work, expression, manifestation, item
• Group 2 entites: person, family (added from FRAR), corporate body
• Group 3 entities: concept, object, event, place

In entity-relationship diagrams the lines indicate where relationships are, and a dashed-line box around a bunch of entities means that the relationship can be with any of the entities inside. So here, a work could be about a person and a corporate body (a biography of Henry Ford and Ford Motors), or about an event and a place (the signing of the Magna Carta and Runnymede), or about a person and a place (any book about someone living somewhere for a year), or an item and an event and a person and a person and a place (Frodo and Sam throwing the Ring into Mount Doom), etc. Mix and match as you need.

The Group 2 entities can be creators and owners and producers of works. If Moreland (person) composes (creates) a symphony (work), and Jenkins (person) publishes (produces) the score (expression) in print (manifestation), and the Warminster Symphony Orchestra (corporate body) performs (realizes) the symphony, and Donners-Brebner Music puts out (produces) a CD (manifestation) of it, and Anne (person) buys (owns) a copy (item) of it to give to her sister Peggy (person) for Christmas (event), then you can see how lots of entities can all relate.

Don’t worry, there won’t be a test. This may come in handy, though.

For the same relationships expressed in a different way, and some new relationships, see what Ian Davis and Richard Newman did:

• Expression of Core FRBR Concepts in RDF
• Expression of Extended FRBR Concepts in RDF

It’s all nicely and readably laid out.

This is background information for the 2006 FRBR Challenge. It may serve as a useful reference. All the lists below are taken from the FRBR Report.

First of all, remember that the Group 1 entities are work, expression, manifestation, and item. Second, remember that a work is realized in an expression, an expression is embodied in a manifestation, and a manifestation is exemplified in an item. Those are the three basic relationships between the four Group 1 entities. To be clear on when two works are related or when something is an expression or manifestation of a given work, look at Barbara Tillett’s family of works diagram, which can be found on page four of the What is FRBR? (2.8 MB PDF) booklet she wrote for the Library of Congress. (If you want to be clear on FRBR in general, that’s a great place to start.)

These relationships can exist between works:

• successor: has a successor / is a successor to (for example, A has a successor B / B is a successor to A)
• supplement: has a supplement / supplements
• complement: has a complement / complements
• summary: has a summary / is a summary of
• adaptation: has an adaptation / is an adaptation of
• transformation: has a transformation / is a transformation of
• imitation: has an imitation / is an imitation of
• whole/part: has part / is part of

There can be these relationships between expressions:

• abridgement: has an abridgement / is an abridgement of
• revision: has a revision / is a revision of
• translation: has a translation / is a translation of
• arrangement (music): has an arrangement / is an arrangement of
• successor: has a successor / is a successor of
• supplement: has a supplement / supplements
• complement: has a complement / complements
• summarization: has a summary / is a summary of
• adaptation: has an adaptation / is an adaptation of
• transformation: has a transformation / is a transformation of
• imitation: has an imitation / is an imitation of
• whole/part: has part / is part of

These relationships can exist between an expression and a work:

• successor: has a successor / is a successor of
• supplement: has a supplement / supplements
• complement: has a complement / complements
• summarization: has a summary / is a summary of
• adaptation: has an adaptation / is an adaptation of
• transformation: has a transformation / is a transformation of
• imitation: has an imitation / is an imitation of

These relationships can exist between manifestations:

• reproduction: has a reproduction / is a reproduction of
• alternate: has an alternate / is an alternate of
• whole/part: has part / is part of

This relationship can exist between a manifestation and an item:

• reproduction: has reproduction / is a reproduction of

These relationships can exist between items:

• reconfiguration: has reconfiguration / is a reconfiguration of
• reproduction: has reproduction / is a reproduction of
• whole/part: has part / is part of

Some relationships depend on whether the two works in question are autonomous or referential. “A referential work is one that is so closely connected to the other work in the relationship that it has little value outside the context of that other work. An autonomous work is one that does not require reference to the other work in the relationship in order to be useful or understood” (FRBR Report section 5.3.1). For example, a referential work can complement a musical work if it’s a cadenza, libretto, or choreography; an autonomous work can complement another work if it’s incidental music (such as for a play) or it’s a musical setting of text.

See the full FRBR Report in PDF form for more details and examples. And remember that if you apply FRBR yourself, you can use whatever relationships you need. There’s no law saying you can only use the ones above. You can make up your own to serve your own purposes.

Here’s Raymond Chandler’s The Big Sleep at LibraryThing, a shared system for cataloguing personal libraries. All of the different editions—manifestations—entered by different users are gathered together as the same work. Note the left-hand side. I think the top cover image is the cover of the most popular manifestation, and the other three are the next three most popular manifestations. Then there are eight links, each for other manifestations. That’s twelve manifestations (if I’m not double-counting duplicates) of one work. Two of the manifestations are translations, so the expression layer is collapsed.

There are details of what LibraryThing considers a work and what makes two things the same work. Those definitions closely match FRBR, but notice the last one: “A Greek edition of Homer is not the same as an English translation. Socially, the former connects you with other Greek scholars, and should recommend other Greek-language works, not the ‘Classics of Western Civilization’ works that the English translation does.” In FRBR terms, they’re the same work. For The Big Sleep, translations are the same work, as we saw above. I’d like to find out more about how Homer’s writings (and other Greek and Roman classics) are separated from other works and how it affects what users see.

This is a very nice piece of work! All of the tags and ratings and comments for all the different manifestations of something such as The Big Sleep come together at the work level, and their power is that much greater because the groups of book owners making them are larger and not divided by trivial differences of ISBNs. If Seamus has the 1970s Vintage edition and Marla has a Penguin edition, their tags and ratings are pooled instead of apparently referring to two completely different books. Users will get much more out of the system now.

When there’s some way to handle an omnibus edition such as the Library of America’s Stories and Early Novels of Chandler, which includes The Big Sleep, it’ll be all the better, though that’ll be even more difficult work. That’s aggregates! This is scheduled as a future improvement, says the page linked above.

There’s no mention of FRBR in the announcement of this and other new features. I wonder how Tim Spalding is doing the work of extracting works from manifestations. Congratulations to him on this stealth FRBRization of a home cataloguing tool. If you’re a LibraryThing user, by all means leave a comment about this if you have one. I think this is an exciting development.

Olivia M.A. Madison, Dean of the Iowa State Univerity Library, has an article in the new Library Resources & Technical Services (50: 1): “Utilizing the FRBR Framework in Designing User-Focused Digital Content and Access Systems.”

ABSTRACT: This paper discusses the rapidly expanding environment of emerging electronic content and the importance of librarians to partner with new research and teaching communities in meeting users’ needs to find, identify, select, and obtain the information and resources they need. The methodology and framework of the International Federation of Library Association and Institutions’ Functional Requirements for Bibliographic Records could serve as a useful tool in building expanded access and content systems.

Madison was chair of the original IFLA study group when it wrote the FRBR report. She recently wrote “The Origins of the IFLA Study on Functional Requirements for Bibliographic Records,” in that Cataloging & Classifcation Quarterly (39: 3/4) special on FRBR last year, which is the thing to read if you want to know how all this got started.

Roy Tennant did a webcast on 17 January 2006 for Infopeople called Leading Edge Technologies for Libraries. He spoke for an hour, and the audio is available (67 MB MP3). The whole talk is worth a listen; I mention it here because he discusses FRBR for several minutes starting at about the 16 minute mark, demonstrating RedLightGreen and OCLC’s Curiouser.

The week after next, on Wednesday 1 March, I’ll announce a contest that will run for two weeks: the 2006 FRBR Challenge. I’ll be posing a task (an enjoyable one, don’t worry) and anyone who enters an answer will have a chance to win a prize. First prize is a Nancy Pearl Librarian Action Figure and a copy of Out On the Rim, by Ross Thomas (40 KB PDF). Second prize is a copy of Gone, No Forwarding by Joe Gores. Thomas and Gores are two of the great crime writers, and two of my all-time favourites. Anyone may enter the Challenge, and winners will be chosen at random.

A fortnight ago I made the rounds of the exhibit floor at the 2006 Ontario Library Association conference and asked some integrated library system/catalogue vendors about FRBR. A fellow at Innovative Interfaces said they’d been doing some work on it but then sort of pulled back, and were waiting, though they’d like to do more work on it if they could find someone to do it with. A woman at SirsiDynix took my card and said she’d have someone get in touch with me to give me an official answer. I’ll let you know when I hear more, from them or Innovative Interfaces.