Identify
By
And
More than six years ago, Helen L. Henderson, Managing Director of Information Power, Ltd., recognized the need for efficient provision of authoritative institutional information in support of the electronic delivery of information. After years of exploring alternative approaches with such organizations as Catchword, the Association of Subscription Agents and OCLC, Helen was persuaded by publishers and standards organizations to bring together a group of interested organizations to propose a new service that would allow authoritative institutional information to be held centrally and to allow that information to be shared reliably and securely. This would reduce the effort of both publishers and users to keep each other up to date with changes in institutional information and credentials needed for authentication. Groups such as COUNTER, ICEDIS and ONIX consider the institutional authority database an essential element to provide consistent and comprehensible measurement of how electronic resources are being used and managed. In July, 2005, Ringgold, Inc. merged with Information Power and is implementing this proposal, in the form of Identify, a service to meet the need for authoritative institutional information. Identify comprises:
Identify's business model allows free access to the database by the users (libraries, institutions, etc.), and for a general look-up service. The revenue to support the service will come from charging vendors, information providers, or other companies wishing to access the database for downloading with a variety of added value services available for subscription fees or one-off charges.
The need to identify institutions is hardly new; taxing authorities have long used identifying numbers to differentiate their tax paying entities. Dunn and Bradstreet established the D-U-N-S Number to improve their financial reporting. D&B reports 66 million currently active companies in their database.
R. R. Bowker initiated and maintains the SAN (Standard Address Number) identifier to signify a specific address of an organization in (or served by) the publishing industry. It is now a USA standard (ANSI/NISO Z39.43-1993) and over 34,000 SAN numbers are active.
The Library of Congress' MARC Code List for Organizations has its roots in over 100 years of organization identification, beginning with Charles Evans' American Bibliography first published in 1903. LC reports that the List contains 27,719 valid identifiers for general use.
We now have an international standard identifier, ISIL, Information and documentation - International Standard Identifier for Libraries and Related Organizations (ISO 15511-2000) - and the Registration Authority started work in 2004. There are 9 countries that have designated registrars (Australia, Germany, Denmark, Egypt, Islamic Republic of Iran, The Netherlands, New Zealand, Norway, Switzerland) and two more are "under registration" (United Kingdom, United States of America).
There are many other institutional identifiers about and one could create still another. However, reviewing the host of institutional identifiers leads us to a different conclusion. We should focus our efforts on improving the communication between publisher, aggregator and institutional consumer, by delivering accurate, inexpensive information about the consumer, including present available identifiers to assist in the identification of the institution.
It should be noted that, in contrast to the international standard metadata for bibliographic data (MARC21), there is no such international standard metadata for organizations. As a practical matter, this means that anyone building a database must define the metadata and the method of representation. It further means that the publishers will have done this same thing and their metadata and their organizational identifications will overlap, but are not likely to be identical. Thus, an Identify organization is defined at an elemental level so that it matches but one organization within each publisher's system. Finally, both organizations and publisher pricing policies are dynamic. As a result, there will be continuous change in the metadata maintained for the institution.
There are over 300 publishers listed in the members lists of ALPSP (the Association of Learned and Society Publishers), SSP (the Society for Scholarly Publishing and STM (the International Association of Scientific, Technical and Medical Publishers). The majority of these publish journals which are available electronically and have hosting and subscription services either in-house or outsourced. Until their journals went electronic the only information they needed about their customers were their mailing addresses. Now they need information about who are they, where they are, what their constituent parts are and exactly what their electronic entitlements are.
The information consumers, and particularly libraries, do certainly have a role in maintaining the institutional information necessary for access, and in many cases they have to register with 100 or more separate services. This means that every time they change any of their credentials (IP address ranges, OpenURL resolvers, etc) they have to contact all the services. There is, therefore, an opportunity to change this relationship from "many to many", to "one to one", by setting up a single source of information on each institution, giving the information needed to identify the institution, its rights, such as subscriptions or licenses, and the correct delivery points, e.g., the institution's IP address ranges.
The needs of the information provider go beyond the atomized information about the individual institution. There are relationships among institutions that must be recognized in order for the provider to recognize the services that are appropriate. For example, the research laboratories of a multi-national corporation may have different levels of access, while subscriptions may be purchased for the joint benefit of members of a consortium who bear no other organizational relationship to each other.
Working with a group of academic publishers, Ringgold has created its Identify database to maintain, search and display all manner of identifiers and all manner of organizational relationships. The database contains authoritative records for institutions in 18 countries (USA, Canada, UK, Ireland, Australia, New Zealand, Netherlands, Finland, Sweden, Norway, Denmark, Germany, Switzerland, Austria, France, Italy, Spain, India, China and Japan), and includes academic, government, corporate and independent organizations. We will continue to maintain and extend this database by providing product and services valuable to publishers and other information providers.
There is a debate to be had over the use of the DOI as an identifier for institutions. Certainly, we could assign DOIs, or we could use DOIs assigned elsewhere. But if they are assigned elsewhere, there will need to be sufficient metadata associated with them to discern where in the network of organizations the individual DOI is applicable. As with any other organization, maintaining the integrity of our information requires that we use an identification number that is under our control; however, that does not reduce the desirability of communicating via the identifiers maintained by other organizations. When we encounter an organization lacking one or more appropriate identifiers, we would be happy to report the institution information to the issuing agency, or to give the institution the links needed to register.
The services offered will provide an opportunity for both information providers and users to reduce the administrative burden of maintaining reliable information relating to the sales, provision and management of online products
The benefits for publishers accessing a central database of accurately defined and uniquely identified institutions include:
The benefits for the user community include:
Ringgold recognizes that the success of Identify will depend upon its accuracy and timeliness. At the same time, we now recognize how much effort is required, not only to create, but also to maintain accurate information about organizations. To accommodate the joint demands of accuracy and timeliness once operational, we will accept immediately changes input by the affected organization, but we will also mark these changes as unverified by Ringgold. Until the changes are verified, the previously verified information will still be available, so that both forms will be displayed. The following metadata is currently available in the database, and Ringgold is actively reviewing needs for all constituencies:
| Institution Identifiers | Ringgold assigned, Government (tax) id, ISIL, LCMOC. D-U-N-S, SAN, |
| Institution Name | Legal name in native language (where appropriate) |
| Alternative Names | Commonly used names, English language name |
| City | City of administrative HQ |
| Region/State | Where appropriate |
| Zipcode | If available |
| Country | Country of administrative HQ |
| URLs | |
| Tier | Carnegie category, JISC band |
| Size | number of staff, students, employees, number of beds, lawyers (as appropriate) |
| Credentials | IP address range, Athens domain, Shibboleth attributes |
| URL resolver | |
| Organization logo | Where applicable |
| Linked Institutions | Name and Nature of Link |
The main services available using the Identify database are:
Librarians and in particular library administrators will be able to access information relating to their institution(s) via an information service (though it is proposed that such services be offered gratis) which would include:
There are a number of other groups which could have additional services or additional options within a service based on their type, for example:
Ringgold's principal concern with Identify is that it be "right", that is, that it be of the highest quality and reliability. As indicated in the Introduction, Helen Henderson has discussed the needs and potential of institution identification with a large number of professionals in the field. She has received preliminary, affirmative responses from those listed below to participate in an Advisory Board for Identify, to help ensure that it meets the requirements of the vendors, information providers and the users:
Librarians:
Publishers:
This is the first demonstration of Identify. We will be presenting it at the Frankfurt Book Fair and at the Charleston Conference.
The Identify demonstration database contains real data, but it is a freestanding database. Before going operational, we will redo the table structures of the database to correspond with our other databases.
We are ready now to sign up a few publishers to move Identify to production status. We expect to wait until after the holidays to begin intensive work with these publishers on their interfaces and workflows, with the expectation of going operational in the period March 31 - June 30. One aspect of this work that will be carefully delineated will be the interaction of customer support. We expect that the publisher will want us to be the first point of contact in the case of service failures. In that case, we will need to be "in the loop" so that we can assure that the service has been reinstated satisfactorily. In the electronic environment, successful customer service is 24x7x365 and we will provide it.
Note: This paper was prepared for the Institutional Identifiers Stakeholder Meeting on 7th October 2005.
