The FRBR Blog

Stefan Gradmann, author of “rdfs:frbr: Towards an Implementation Model for Library Catalogs Using Semantic Web Technology” which was in the special FRBR issue of Cataloging & Classification Quarterly (39:3/4, 2005), has posted it to the E-LIS e-prints archive. (You’ll need to register to read it, which I think is new, but registration is free and easy.)

ABSTRACT: The paper sets out from a few basic observations (bibliographic information is still mostly part of the ‘hidden Web,’ library automation methods still have a low WWW-transparency, and take-up of FRBR has been rather slow) and continues taking a closer look at Semantic Web technology components. This results in a proposal for implementing FRBR as RDF-Schema and of RDF-based library catalogues built on such an approach. The contribution concludes with a discussion of selected strategic benefits resulting from such an approach.

If you’re new to this then you’ll also want to see Ian Davis and Richard Newman did about making an RDF schema for FRBR. Check out the other blog entries in the Semantic Web category, too.

(Seen on Catalogablog.)

Amy Hart wrote an interesting introduction to FRBR, “FRBR and School Libraries,” in the August/September 2006 issue of Library Media Connection. (I found it through my public library’s web site, which offers access to EBSCO’s MasterFile Premier.)

Not too many years ago while our shelf lists for retro-conversion to online MARC records, a colleague at Lexington (Massachusetts) Public Schools wondered why we had to have separate records for each of the versions of the same book. At their library media center, they added additional or replacement copies of a title to the original shelf list card, making notes at the copy level if there was different ISBN or publisher information. The library media specialist voiced the opinion that it made sense to list all the versions of Charlotte’s Web on one card since kids just want the story and don’t care whether it’s a republication, paperback, hardback, or reprint. Little did we know that while we were busy assigning the different versions on our single cards to separate MARC records, the International Federation of Library Associations and Institutions (IFLA) was already working on a plan to reunite them.

That’s a useful example of how different kinds of users need different sorts of displays. If that school library were to be fully FRBRized now, I wonder what kind of online catalogue would work best for the current students, who probably browse the web and play video games a lot.

A quick quote from some mail that Michael Gorman had someone forward to the RDA mailing list: “At first glance one of the very few things that I liked about the chaotic mess offered as the proposed Rules for descriptive cataloguing (RDA) was the addition of ‘families’ when ‘corporate bodies’ are mentioned.” (People and corporate bodies are Group 2 entities in FRBR, and families are added by FRAR, if I remember right.) He proposes that they both be rolled up into ‘collective entities.’

The 2006 IFLA conference is happening right now in Seoul! The Review Group (the overall FRBR committee), its working groups on aggregates and expressions. and the FRASAR and FRAR working groups are all meeting. I’ll let you know as soon as I find out what happened.

The 2006 FRBR Challenge ended on 15 March. I think enough time has passed that the excitement has died down and we can consider the Challenge soberly.

Here’s what I said when I announced it:

The challenge is: How well does FRBR handle Middle-Earth?

People are using FRBR, some on a small scale, some on a large scale, some by automating the extraction and matching of works from a set of manifestations, some by having people do the work manually, most by a mix of both. They’re all showing that FRBR works and that it helps users. They’re also turning up some problems, points that need clarification, grey areas, and places where the model falls a bit short and needs something added. That’s to be expected.

Applying FRBR to J.R.R. Tolkien’s Middle-Earth (The Hobbit, The Lord of the Rings, and everything else) will test the model and give us some ideas about what’s easy to do and what’s hard to do, what’s clearly explained in FRBR and what needs more detail, what FRBR already handles and where it’s lacking, and how to deal with really complicated situations involving digital copies or multiple works in one work. I really do mean everything else: movies, Sindarin dictionaries, the whole megillah.

The Rules

So here’s what to do: think of something LOTR-related, and, as best you can, describe it as a work, expression, manifestation, and item, and, if you can, explain a relationship it has to a book Tolkien wrote or something inspired by Tolkien’s work. That’s all. Use official FRBR terminology if you want, but you don’t have to. If you can’t think of what the expression is, or the work, just say so and maybe someone else will have a stab at it.

Response from readers was better than I’d hoped for: four people submitted entries and they all won a prize. You’re sorry now you didn’t enter, aren’t you? But you can try next year.

The Entries

I’ll leave out most of the details here and just concentrate on the works.

• Work: The Hobbit by J.R.R Tolkien
  • Expression: The Annotated Hobbit, or There and Back Again, a revised and expanded edition annotated by Douglas A. Anderson.
• Work: The Fellowship of the Ring
  • Expression: Either Tolkien’s original text or his revised text
    • Manifestation: Unwin Paperback’s second edition (unknown reprinting) of 1966 (ISBN 004823155X)
  • Expression: 1993? text edited by Douglas A. Anderson
    • Manifestation: Harper Collins, ISBN 007123825
    • Manifestation: Collins Modern Classics, ISBN 00712970X
• Work: The Fellowship of the Ring, a game from Iron Crown Enterprises.
• Work: The Lord of the Rings: The Fellowship of the Ring, 2001 movie directed by Peter Jackson
• Work: The Lord of the Rings: The Fellowship of the Ring: Gandalf the Wizard Glass Goblet (from Burger King in 2001)
• Work: A Guide to Middle-Earth by Robert Foster
  • Expression: Author’s edited text, 1971
• Work: The Complete Guide to Middle-Earth: From The Hobbit to The Silmarillion, by Robert Foster
  • Expression: Author’s edited text, 1978 (substantially revised version of the above, enough to make it count as a new work)
• Work: A clip-on costume Gandalf beard.
• Work: Nightfall in Middle-Earth by Blind Guardian (a German power metal band), based on The Silmarillion
• Work: The Return of the King, by J.R.R. Tolkien.
  • Expression: Tolkien’s abridgement (unknown date), read by Tolkien himself
• Work: The Atlas of Middle-Earth, by Karen Wynn Fonstad
• Work: Quenya Course by H.K. Fauskanger
• Work: Bored of the Rings, by Henry N. Beard and Douglas C. Kenney

The Relationships

Here’s how all the works are related. The names of the relationships are drawn from section 5 of FRBR Final Report. (I didn’t get into what Ian Davis and Richard Newman defined in their RDR specification of FRBR.)

• The Hobbit to The Lord of the Rings: has successor?
• The Lord of the Rings to The Silmarillion: has successor?
• The Hobbit to The Silmarillion: has successor?
• The Lord of the Rings to The Fellowship of the Rings, The Two Towers, The Return of the King: has parts
• The Lord of the Rings and The Fellowship of the Ring to The Fellowship of the Ring (game, which involves more than just the events of that one book): has transformation
• The Fellowship of the Ring and The Two Towers to The Fellowship of the Ring (Peter Jackson’s movie, which ends with the start of the second book): has transformation?
• The Fellowship of the Ring (Peter Jackson’s movie) to the Burger King goblet: ? (has commercial tie-in)
• The Lord of the Rings to a clip-on costume Gandalf beard: ? (has commercial tie-in)
• Nightfall in Middle Earth to The Silmarillion: is complement (autonomous musical work)
• A Guide to Middle Earth to The Lord of the Rings, The Silmarillion, The Hobbit: is supplement (referential work)
• The Complete Guide to Middle Earth to The Lord of the Rings, The Silmarillion, The Hobbit: is supplement (referential work)
• A Guide to Middle Earth to The Complete Guide to Middle Earth: has successor
• The Atlas of Middle Earth to The Lord of the Rings, The Silmarillion, The Hobbit: is supplement (referential work)
• Bored of the Rings to The Lord of the Rings: is imitation (parody)

So, what does that tell us? The entities (work, expression, manifestation, item) are clear enough. What struck me is the difficulty of defining all the relations. Those relations are important. Users will want to know them when they’re searching and browsing.

I’m not sure of the proper term to describe how Tolkien’s books (The Hobbit, The Silmarillion, etc.) all relate. It is not clear how to express the relation of a work to its commercial tie-ins. “Has transformation” covers a vast range of possibilities and would need to be further defined. The list of relations in the Final Report is just a start, I think. If FRBR is to go into widespread use, we’ll want a way to standardize, or at least share, what we call all those relations. Same for work-to-expression relations, expression-to-expression, expression-to-manifestation, and so on, including the relations to people and subjects.

We need to consider both end users and the metadata people when we think about the relations. The metadata people will want a lot of granularity and precision. End users may not — by default a display probably wouldn’t show them a lot of detail, but they might want to see more. How much detail is good? How could the data be stored? How could it be shared? How much can we pull out of, or infer, from existing MARC records?

Conclusion

FRBR is well up to the challenge of mapping out the complicated bibliographic universe that is Middle-Earth. There are some points that need clarification or definition. If Tolkien’s complete writings were FRBRized there would be a lot more questions, probably including many about the expression level. (There’s a working group taking a closer look at expressions, but I have no idea when it will publish something.)

I propose a full-scale test of FRBR by FRBRizing the complete Middle-Earth works of J.R.R. Tolkien, or, for an easier test, FRBRizing the complete works of J.K. Rowling. Rowling’s works are sufficiently well-known and well-documented, even the translations, that it would be a reasonable class project or master’s thesis.

I mentioned Wikicat in June because it uses FRBR (see the Meta-Wiki page about Wikicat). Richard Wallis of Talis took a more in-depth look at it yesterday in a blog entry called Wikicat:

In simple terms the Wikicat project is attempting to assemble a bibliographic database [yes another one] of all the bibliographic works cited in Wikimedia pages.

… Unlike Wikipedia where some concerns about data quality are overridden by the value of it’s totally participative nature, the Wikicat team are clearly aware that the value of a bibliographic database is directly connected to the quality, consistency and therefore authority of the data that it holds. For this reason, the establishing of cataloguing rules and training for potential editors overseen by the WikiProject Librarians is already well detailed in the project operational stages roadmap.

A bilingual (English and Italian) blog entry from Bonaria Biancu, the Geek Librarian: FRBR secondo OCLC e Library Thing. It’s about OCLC’s xISBN and LibraryThing’s thingISBN.

It’s probably no news to you that worldcat.org is now open: all of the OCLC’s enormous union catalogue, WorldCat, is available for your searching. It has lots of features, so explore it if you haven’t already. The editions/manifestations groupings are there as before: look at The Three Musketeers (the 1991 Oxford World’s Classics manifestation, ISBN 0192827510) and follow the “Editions” tab to see other manifestations of the same work. If you browse around or search you’ll see faceted search refinements on the left-hand side and language facets in a top navigation bar. A fantastic resource. Congratulations to OCLC for their work on this.

Here’s something interesting: Deidre Kiorgaard and Ebe Kartus, two Australian librarians, did a presentation on 8 February 2006 at the VALA 2006 conference: “A Rose By Any Other Name?: From AACR2 to Resource Description and Access.” I just found out about it.

• Paper: A Rose By Any Other Name? (44 KB PDF)
• Audio of presentation (Ebe Kartus talking): A Rose By Any Other Name? (6.6 MB MP3)

It’s a good overview of what’s going on with AACR2 and RDA and who’s doing what. FRBR and FRAR are mentioned.

Found through Bronwyn’s Library Blog.

From Introducing the thingTitle API, from LibraryThing’s Thing-ology blog:

Announcing another simple LibraryThing API. Feed it a title and it will return all the ISBNs from the most likely LibraryThing “work,” the LibraryThing title and a link to the LibraryThing work page.

It’s simple to use. If you look at http://www.librarything.com/api/thingTitle/The%20Three%20Musketeers (notice title in URL) you’ll get some XML back including the URL of the work LibraryThing thinks it is and a list of ISBNs of manifestations it thinks go with the work.