Standards, Identifiers, and all that Good Stuff

LITA Standards IG, Saturday, June 24, 2006

Managing Identifiers

Moderator, Pat Stevens

Pat told a story about attending an ISO meeting in Asia where one topic was the international trade in ring tones. She pointed out that public identifiers are critical to link heterogeneous services, they require support—technical and social infrastructure. The presenters were asked to talk about what went on behind the scenes to support identifiers.

Regina Reynolds, US ISSN Center, Library of Congress

“The ‘Business’ of ISSN: What Does it Take?”

The center has been around since the early ‘70s. The ISSN is undergoing a major revision, has 80 centers all over the world, the International center in Paris. Governed by a board and has a User Group which had its first meeting in 2005. New centers need training and documentation. A unique key title and metadata record backs up every ISSN assignment.

Decisions are made at the annual meeting. One example was a recent decision not to assign personal blogs ISSNs (for which Regina was soundly vilified by at least one blogger). ISSN assignment is free, but there is an enormous cost to maintain the system, including for title changes and various versions. Budget is 1.5 million euros/year (55-60% is salaries). Revenue comes from country member dues.

ISSN database now has 1,252,191 records as of June 18, growing at the rate of over 50-60,000/year and experiences many maintenance transactions. Access to these records is via the ISSN portal.

Assignments in the US are about 6000/year—is low because many publishers don’t know that various versions/formats need separate ISSNs. Regina hopes to put effort into publisher education to make that clearer. Other challenges: costs/staffing, globalization (ISSN is country-based), “vanity” ISSN seekers, new formats and new technology (hand-held devices are an issue coming up).

Future directions: new products and distribution services, broader coverage, increased interoperability with other standards & services, and increased automation.

Brian Green, International ISBN Agency

“More than you probably want to know about ISBN and other identifiers”

ISBN system devised in the late 1960s, ISO ISBN standard (ISO 2108) first published in 1972, last revised in 1992, then most recently May 2005. UPC introduced a year later. Universally adopted as the key identifier for books. In 1980 the EAN-13 barcode system, based on country prefixes, began to take off. To accommodate books, a new country, Bookland, was created, with its own country prefix. The ISBN community is now discussing incorporation of ISBN into RFID tags.

The ISBN effort was run by the Berlin State Library with no formal agency or governance from 1972-2006. The International ISBN Agency, Ltd. was formed in 2005: not for profit, members are all national agencies, each has one vote (160 of them). Unlike the ISSN is very decentralized—it assigns a group identifier, variable based on the size of the publishing industry in a particular country. A local agency then assigns publisher identifiers, based on number of books anticipated, and the publisher assigns the product portion.

Agencies and publishers in a database available on the ISBN website. There is now an ISBN metadata set, based on a simple ONIX compliant minimum set: ISBN, product form, title, contributors, series, edition, language, publisher, imprint, country, date. Agencies may charge “reasonable” fees for for assignments of ISBN, but some argument about whether the fees are really reasonable. From 2005 national agencies pay membership fees based on GNP and publishing turnover. These funds are used for training, software, and other management tasks.

ISBN is only for books, and manifestations thereof. May be an electronic publication on physical carriers or online (ebooks). A separate identifier is required for each electronic version separately traded. As with ISSN is a “supply chain identifier.” ISBNs can be assigned to parts of books if traded separately (an increasing issue).

The move to 13 digits coming up is very complicated. Standard management tools are now under development, though at present each country has their own system and software. Green seems to be sanguine about the ISBN is meeting and will continue to meet the needs of both libraries and the book trade, including in the digital environment and with increasing issues of granularity. There is an urgent need for interoperability, both horizontal and vertical—accommodating different media, fuller metadata for different purposes, etc.

Green is chair of ISO TC46 SC9, which is the body managing the development of many of the relevant identifier, and the agencies for many of these identifiers are increasingly working together to address issues of interoperability. They are working on use cases and looking at “hub and spoke” metadata mappings to a generic indecs-like generic data dictionary. The use cases are on the ISBN website and Green showed an example of one. The committee has an Identifiers Interoperability Group, which is looking at areas where new identifiers may be needed. One example is the proposed work towards an ISPI (International Standard Party Identifer), an image identifier and license identifiers. Green suggested two websites: http://www.isbn-international.org and http://www.lac-bac.gc.ca/iso/te46sc9/ and cites also Norman Paskin’s article in DLIB (http://www.dlib.org/dlib/june03/paskin/06paskin.html) as a good resource.

Chuck Kosher, CrossRef

CrossRef is more a consumer of standards than a standards agency, and talked about what it takes to keep an actionable identifier alive. He began by talking about the DOI and how it’s used. In this world a publisher registers a DOI with CrossRef, and they use assigned prefixes. Someone else who knows about this object comes to CrossRef to find out where it lives. They can then build a link on their page and rely on it to be actionable by users.

Kosher showed a complex diagram (barely readable in my back-of-the-room location), that it would have been nice to see up close, but it did give a flavor of the task. The community that makes this happen was also described (bottom-up):

CNRI: develops and maintains the hardware to run the Handle system.
DOI: maintains the standard, and works with the community of users
CrossRef: provides the linking services and metadata lookup services to support it. maintains the integrity of the services (metadata and link quality)

A lot of this effort is underwritten by the large commercial publishers. Between 10-13 million people a month are finding articles using this service, according to Kosher (no stats on people who don’t find what they’re looking for).

Support structures needed for:
Data construction
Technical issues and education around them
Metadata issues (“semantics are MUCH harder than syntax”)

Also needed:
Staff to keep the systems running
Creative people to move things forward, add new features, maintain service quality

Question: Since personal publishing is getting more prevalent, and these models are run on national agency models, will they scale enough for these new “publishers?”
Answer (Green): Agencies are often free to deviate from the norm, and some are willing to venture into this area especially if these small people are willing to pay. In smaller countries particularly the agencies are not set up to accommodate this. (Reynolds) These people are a problem in a “free” system. (Kosher) CrossRef relies on annual membership fee, then charges a dollar for a DOI. They use “managing agents” where larger publishers stand in for smaller ones, for a fee. (Didn’t really answer the question).

Question: What’s the difference between the DOI and the Handle System? (Kosher) Not much difference, but DOI is more of a community, with more support and social structure. Handle System is more the technology. If you see a “ten-dot” prefix, it’s DOI; if not is probably just Handle.

Question: What happens when a publisher decides to take something down? (Kosher) There’s been very little of that, in most cases a purchaser of assets takes over the responsibility, but they watch out for dead stuff.

Question: What was ESPN and what happened to it? (Green and Reynolds) This was an entrepreneur who seemed to be confusing people, and he was persuaded to change his M.O. and revise his site. (Reynolds) “I wanted to send the bloggers over there!”

Posted by Diane Hillmann (LITA Standards Coordinator)