The FRBR Blog

[Warning: If you follow other technical and cataloguing library blogs, you probably already know about this. If you don't, start following Planet Code4Lib, where you'll see at least three other blogs posting this announcement.]

xISBN is the OCLC service that accepts an ISBN and returns a list of ISBNs that represent other manifestations of that same work. It’s simple and extremely useful.

They’re doing some work on it. Here’s an announcement Eric Hellman sent to the code4lib mailing list yesterday (except that I corrected 14 to 13):

As many of you are aware, the xISBN service, developed by OCLC’s Office of Research, has been running in an experimental, semi-supported mode, and it has proved quite popular. Last year, OCLC charged the OCLC Openly Informatics Division with the task of making it a fully supported WorldCat service.

At about 4PM EST on Tuesday, February 13, a switch will be flipped, and traffic aimed at the experimental version of xISBN will begin to be routed to a replacement xISBN service supported by the Openly Informatics Division of OCLC. Any application that follows http redirects- this should be most xISBN client applications- will continue to work without needing changes. The timing of this switch has been dictated by the decommissioning of a server, and we apologize if this short notice seriously impacts anyone.

After the switch, the traffic currently sent to http://labs.oclc.org/xisbn/[ISBN] will be redirected to http://old-xisbn.oclc.org/webservices/xisbn/[ISBN]. This service will respond in almost exactly the same way that the research version has responded; you can change your applications to use the replacement address effective immediately. Of immediate benefit to all users of xISBN is be the drastically improved currency and frequent updates of the xISBN data set.

As you might guess from the replacement system host name, there will soon be a “new” version of the xISBN service. Xiaoming Liu, who has been working on xISBN for 3 months, will unveil the “WorldCat xISBN Service” at the Code4Lib conference at the end of the month.

There is a small difference in the behavior of the replacement service. If you send the replacement service a 13 digit ISBN, the entire result set will be returned with 13 digits.

If you expect your xISBN client service to use more than 1000 queries per day, please let us know (xisbn-support@oclc.org), as the traffic control systems have also changed.

To make sure that you are alerted of all of the coming changes surrounding xISBN, please make sure to sign up for the XIDENTIFIER-L listserv. Sign up at http://listserv.oclc.org/scripts/wa.exe?A0=XIDENTIFIER-L.

Related blog posts: Thom Hickey, also of OCLC, comments, and Talis’s Panlibus blog chips in.

Following up on her UCLA Film and Television Archive OPAC announcement that I posted here on Wednesday (which includes some links to follow to learn more about her FRBR-related work), today Martha Yee answers the Four FRBR Questions!

When did you first hear about FRBR?

In the course of FRBR development (1992-1998), in order to help in filling out the tables in the back of FRBR that link FRBR entities to particular elements of the bibliographic description, the Study Group sent out a questionnaire to people in the field, and I was one of the people who was asked to fill out a questionnaire, in my case concerning moving image material. The definitions of the entities were still not clear, and I recall writing to ask for clarification before sending in my filled out questionnaire, and feeling afterwards that I had not understood the intended definition of “manifestation,” so had sent in erroneous information! But I can’t remember the year that questionnaire was sent out.

What’s your involvement with it now?

I try to follow developments from a distance and comment upon suggested changes when asked to do so, or given the opportunity to do so.

What’s one thing you think the FRBR world needs most?

Ability to be involved in the design of new and better catalog interface software from the ground up, as well as a world-wide effort to rethink the process of shared cataloging in this new era of the Internet. Could we design excellent standard catalog interface software to be used by everyone everywhere to access a single virtual catalog, the maintenance of which is shared by all catalogers everywhere? Much of what blocks our ability to implement FRBR effectively is essentially our current method of shared cataloging, which involves sharing atomized manifestation records in an environment of thousands of separate catalogs, with inefficient methods of coordinating what we call the FRBR entities across all catalogs.

What’s your one-line non-librarian description of FRBR?

A model for transforming records for the books published by publishers into displays of the entities that users actually seek: 1) works, which may have been published thousands of times under many different titles associated with many different variants of their author’s names; 2) persons and corporate bodies, which may be known by many different variants of their names; and 3) concepts discussed in works, which may have many synonyms, and which may be known by homonyms that also represent completely different concepts.

My thanks to Martha Yee for taking the time to answer the questions. She’s the fourth person in the Four FRBR Questions series; the other three are:

• August 2005: Patrick Le Boeuf
• August 2005: Pat Riva
• January 2006: Barbara Tillett

Kristin Antelman works at the North Carolina State University library and has written about FRBR and posted comments right here pon this blog. In the spring of 2006 she was interviewed for a podcast series from the Coalition for Networked Information. You can download the interview directly (15 MB MP3). It’s about the debut of the spiffy NCSU library catalogue, which is now slightly old news, but I mention it because at about the 12:00 minute mark Antelman mentions FRBR and their plans to implement it. The whole interview is interesting and worth a listen.

Come back tomorrow to see Martha Yee do the Four FRBR Questions!

If you read this blog, and you do, because you are, you know about Martha Yee. She’s Cataloging Supervisor at the University of California at Los Angeles’s Film and Television Archive. It’s the kind of place where, for example, as happens Friday, Shirley MacLaine, Audrey Wilder, and Curtis Hanson come by for a discussion and a showing of The Apartment, the great Billy Wilder movie with MacLaine, Jack Lemmon, and Fred MacMurray. As a film, it’s one of the best ever, classic-wise.

She’s often mentioned on this blog. Here are a few relevant posts:

• January, she spoke at the FRBR implementers meeting at the ALA conference.
• September 2006, she spoke at the Beyond the OPAC conference in Australia.
• October 2006, she was announced as contributing a chapter to Arlene Taylor’s upcoming book on FRBR (due out this spring).
• July 2005, her paper FRBRization: A Method for Turning Online Public Finding Lists into Online Public Catalogs came out.

Yesterday she sent out this announcement:

We (UCLA Film & Television Archive) released our new OPAC interface today, at: http://cinema.library.ucla.edu

Try out the pre-existing works search for an example of a search of superworks using existing records (main entry and uniform titles) and existing catalog technology (Voyager); a search on shakespeare, or a search on hamlet, or a search on maugham will give you a good demonstration. Note the value of providing a keyword in heading search with a display of matched headings, so that users need not know the entry terms in our headings in order to find the entities they seek.

Unfortunately, Voyager will not allow us to build an A-Z title index using both title added entries in bibliographic records and title cross references in authority records. For that reason work searchers in our catalog must always do two searches (title and title variants) to be sure we do not have the moving image work they are seeking under title. TAOS did allow us to build such an index, so for this particular feature, Voyager represents a step backwards for us.

Unfortunately, Voyager will also not allow a true main entry sort of bibliographic records retrieved (130 if present, if not, 245), and the sort for the left to right title index (our default search) dedupes to the first title of any kind in the alphabet on any given bibliographic record, so displays of multiple records can easily fail to display all the expressions of a work together.

We still have a long way to go on the road made by walking!

We create work-based authority records (you can see an example if you do a title variants search on Raise the red lantern, and then click on the “notes” button–Voyager won’t let us display the whole authority record, but at least we can display the public notes). Our bibliographic records are expression-based, and we attach multiple MARC 21 holdings records for multiple manifestations (differences in format and distribution information without differences in underlying content). Even though this is “cheating” in the current shared cataloging environment, it was absolutely imperative to do it in a preserving film archive. One of our preserved films, Becky Sharp, is a good example of why this is imperative!

Pop open a fresh tab or window and try out the searches she suggests.

This came out last year but I just found out about it thanks to someone tagging it at del.icio.us: A Study on the Application Method of the Functional Requirements for Bibliographic Records (FRBR) to the Online Public Access Catalog (OPAC) in Korean libraries, by Jane Cho (Library Collections, Acquisitions, and Technical Services 30:3-4, September-December 2006).

ABSTRACT: The work-set algorithm to automatically convert a bibliographic database into a Functional Requirements for Bibliographic Records (FRBR) structure has been developed and operated. However, in case the authority control is not properly performed as in Korean catalogs, it is impossible to FRBRize using only the automechanism. In such a case, a visual check must accompany the process; unfortunately, this requires extensive time and cost. Thus, in this case, it is desirable to cluster work-sets for a large-scale bibliographic database like a union catalog and to share the resulting work-sets with local libraries. To that end, this paper presents a method for extracting work-sets from the Korean union catalog using a semi-automatic mechanism. Additionally, it proposes methods to allow local libraries to apply the FRBR to their own OPAC using work-sets that result from the union catalog.

As I’m sure you recall, back in June 2006 in Eprints Application Profile, I informed you that Julie Allinson (whose last name I regret to say I misspelled) had dropped me a note about the Eprints Application Profile, which described a plan to develop “a Dublin Core Application Profile for describing scholarly publications (eprints) held in institutional repositories.”

Now it’s finished and Allinson, Pete Johnston, and Andy Powell have published a paper about it: A Dublin Core Application Profile for Scholarly Works (Ariadne 50 (January 2007)).

In May 2006, the Joint Information Systems Committee (JISC) approached UKOLN and the Eduserv Foundation to collaborate on the development of a metadata specification for describing eprints (alternatively referred to as scholarly works, research papers or scholarly research texts). A Dublin Core (DC) application profile was chosen as the basis of the specification given the widespread use of DC in existing repositories, the flexibility and extensibility of the DCMI Abstract Model and its compatibility with the Semantic Web. The main driver for this work was the establishment of a three-year project to aggregate content from repositories and offer cross-searching and other added-value services.

… The application model for scholarly publications presented here is based on the Functional Requirements for Bibliographic Records (FRBR).

It’s fairly short, so there’s no reason you can’t read it all. It’s interesting to see how they’ve adapted FRBR for their own specific purposes, what relationships they define between the entities, and what attributes the entities have. Nice work!

Some comments on it:

• Lorcan Dempsey’s Metadata Complexes
• Dublin Core and Scholarly Publications on the eFoundations blog, with a comment by Bruce D’Arcus.
• Abigail Bordeaux (which I noted last Friday) and Leslie Johnston both saw Allinson talk about the application profile at the Open Repositories conference a few days ago.

Remember, this blog covers things half a year before you’ll read about them elsewhere. Don’t be left behind. Read the FRBR Blog.

• Judith E. Bush’s 80 millionth OCLC Number and a Personal Refresher on FRBR mentions sheet music and a MIDI version of a musical work as different expressions. (She wonders if she’s right about that, and she is, as long as it neither has been put into physical form, in which case they become manifestations and items.)
• Mark Lindner’s And On a Related Topic … Fun in the Classroom describes what sounds like a great class discussion about unexpressed works and unrealized expressions. Then he gets into some thoughts about whether FRBR can handle imaginary works (such as Camel Ride to the Tomb by X. Trapnel or Match Me Such Marvel by St. John Clarke). On top of that, I’d add the many worlds interpretation of quantum mechanics and ask how FRBR can handle works etc. that exist only in parallel universes (which would include all those imaginary works). (I think all cataloguing and classification systems should be able to handle things made by aliens or in parallel worlds. “The library is a growing organism.”)
• Deborah Kaplan’s LibraryThing and FRBR concerns, as you would expect, LibraryThing and FRBR.
• Abigail Bordeaux was at the Open Repositories conference and FRBR came up there, as she reports in Days 3 and 4: Sunny and Sixty. People said “copy” instead of “item.”
• A few weeks ago I mentioned Frédérick Giasson’s Music Ontology and now he’s made version 1.02 available.

OCLC’s Thom Hickey saw Jonathan Rochkind’s report on the FRBR implementers meeting at the recent ALA conference and followed up on his own blog with more explanation of (to quote him) the differences of the groupings used for WorldCat.org, xISBN, and FictionFinder.