Institutional registries and identifiers: comparison of current models
Last month, OCLC announced the launch of its WorldCat registry, intended as a directory wherein libraries could "manage and share data that define their organizations - such as institution type, location, URLs for electronic services, circulation statistics, and population served"1 - a very similar mission, on the face of it, to the other institutional registries in existence, all of which can usefully claim to provide publishers with a single source of reliable institutional metadata, including marketing information about libraries' holdings, budgets and so forth. In this introductory article, we overview the key players and identify some of the differences between them.
OCLC WorldCat Registry
Whilst OCLC's registry claims to be user-focussed (helping searchers to find their nearest library), it is also the latest to be developed with the intention of saving libraries from having to register and update the same information in multiple locations. The hope is that libraries will deposit and maintain information about their library, ranging from an institution's name and type to its branch relationships and consortial memberships. Use of the data is restricted to WorldCat affiliates, eligibility for which status is determined by OCLC. Sharing of data is currently via HTTP but OCLC will provide Web services in due course.
Pros: free to libraries; pre-populated with strong US coverage; institution-type agnostic; privacy and opt-out options for institutions wishing to hide data; forthcoming RSS feeds to alert affiliates to data updates; support for multiple identifiers.
Cons: US-centric focus limits current value to global affiliates2; records have not, at this point, been de-duped3; libraries must maintain data - and, ironically, maintain it separately for other OCLC services4; likely inaccuracies and staleness of self-maintained data; opt out/privacy options means dataset available for affiliate usage will be incomplete5; limited hierarchical capabilities; limited support for affiliates (e.g. for matching data); unspecified licensing costs for affiliates.
In summary: some nice introductory features from OCLC, but experience elsewhere suggests it is naive to expect that the customer (libraries) will clean up the data on behalf of the provider.
Ringgold Identify
Ringgold's institutional registry has a very different perspective; its initial focus is not libraries and their users (although there are user benefits), but those in the publishing supply chain for whom a common identifier could make interaction more efficient and effective. The Ringgold database is centrally maintained by a staff of 25 international researchers, and data can be licensed by publishers and other supply chain intermediaries. Ringgold's auditing services help publishers clean up their own databases and accurately assign identifiers to the correct institutions - along with parent identifiers to indicate that institution's hierarchical relationships with other institutions. To date, the registry includes details of 65,000 institutions in 100 countries. As a participant in the Journal Supply Chain Efficiency Improvement project, Ringgold is actively working to ascertain the benefits and implications of using institutional identifiers to increase efficiency.
Pros: centrally-maintained, ensuring accuracy, uniqueness and availability of data; "vendor neutral"; global focus; support for multiple identifiers; hierarchical capabilities; focus on publisher supply chain needs; supporting services to help publisher implement and maximise value of identifiers.
Cons: less well known in library market - and not focussed on making libraries' lives easier; additional cost in the publishing supply chain - to solve a problem that publishers are less aware of/affected by; publisher/intermediary uptake required to secure future of a standard identifier.
In summary: with strong support from organisations throughout the publishing supply chain, Ringgold's identifier is already becoming a de facto standard, with support from standards bodies like NISO and EDItEUR.
Atypon
Announced the launch of their "AIR" (Atypon Institutional Registry) in 2005, intended to contain a limited subset of information for each institution, including the IP address range, alongside an institutional DOI (digital object identifier - one of the objections to the project was its characterisation of libraries as "digital objects"!). There were also concerns about the impartiality of Atypon as a provider of this kind of service within the e-content sector 6. The company proposed to provide the supporting infrastructure for the project at no charge to the community it would serve, but has made no further announcements about the project.
In addition, e-content providers like Ingenta maintain registration databases to support access to the content they provide. IngentaConnect's database contains 25,000 library registrations and was pioneering in its allocation and use of identifiers (known as CIDs) to simplify transactions with third parties such as subscription agents - these identifiers are still used by our agent partners and are critical to the process by which they automatically activate subscription access on behalf of our mutual library customers. The IngentaConnect interface allows libraries to update any of the details associated with their registration, such as IP ranges and mailing details for their institution - processes which the MetaPress system, for example, cannot handle through a web interface (the library must contact MetaPress by email to make any changes). Another key benefit of IngentaConnect's system is that an institution need have only one registration to use the more than 30,000 publications available from the IngentaConnect website - whereas the HighWire system, for example, requires libraries to register the same information separately for each publisher's journal(s).
Where can I find out more?
Henderson, Helen. "Institutional identifiers: how do we know who we are?" The Charleston Advisor, volume 7, number 3, January 2006. http://www.charlestonco.com/features.cfm?id=195&type=fr
Henderson, Helen. "Institutional identifiers: the Supply Chain pilot to date". The Charleston Advisor, volume 8, number 3, January 2007. http://www.charlestonco.com/features.cfm?id=214&type=fr
Footnotes
1 OCLC Press Release: WorldCat Registry offers management of organizational data, DUBLIN, Ohio, USA, 26 February 2007.
2 http://www.smartech.gatech.edu/bitstream/1853/13633/1/330-thurs-2_05.pdf, accessed 20 Mar 07
3 http://walt.lishost.org/?p=487; http://www.worldcat.org/webservices/registry/xsl/faq#regfaq-duplicate, accessed 20 Mar 07
4 http://www.worldcat.org/webservices/registry/xsl/faq#regfaq-IP, accessed 20 Mar 07
5 http://www.worldcat.org/webservices/registry/xsl/faq#regfaq-dontinclude, accessed 20 Mar 07
6 http://www.diglib.org/architectures/instreg/InstRegMinutes.pdf, accessed 20 Mar 07