Evolving the Resolving: OpenURL

OpenURLs came about because of the proliferation of full texgt data from a variety of sources, particularly databases compiled by aggregators. Since each aggregator has a different database format, different field labels, etc., and since libraries usually subscribe to a number of these services, it’s impossible to specify a single URL to locate an article. Two components need to be integrated into a system: a link resolver, which converts links to articles retrieved from databases, and a knowledge base of the journals that are covered by the databases to which a library subscribes. In addition, if an article is available from more than one source, a library may wish to specify which source to access (the “appropriate copy” problem). This excellent session provided an overview of ways to solve the link resolution and appropriate copy problems.

Nettie Lagace, from ExLibris, said that users want to locate information quickly, but it is usually distributed among several sources. Information providers, not librarians control the links to the information, which leads to high maintenance of proprietary solutions and a less than optimal usage of expensive collections. So users are not being well served. Consider the process for finding an article copy:
1. Access a database and find an article.
2. Write down the citation.
3. Make sure that journal titles are correctly spelled out (not abbreviated).
4. Find out whether the full text is available electronically or only in print.
5. If it’s available electronically, access the database with the full text (which may be different from the first database), search for the article, and print out the full text (which may not include the images, etc.).
6. If it’s available only in print, submit an ILL request and wait 3-4 weeks for the copy to arrive. (The only response I can make to this is “Ugh!”)

The OpenURL standard sets specifications for interoperability and puts control of the links into the hands of the librarian. It makes retrieving from a full text database transparent to users and helps solve the appropriate copy problem. And libraries can customize the interface to conform to their library’s policies and services. A recent development is that Google Scholar can be enabled to display Open URLs to link resolvers, and some full text journals have converted the references to the articles they publish to the Open URL standard, all of which is making finding information much easier for users.

Three librarians (Rafall Kasproski, University of Houston; Cindi Trainor, Claremont Colleges, and Judy Wulff, University of Louisville) presented case studies of their implementation of Open URLs and link resolvers.

Kasproski stressed that it is important to go through a formal selection process and to involve the library community. The entire community should be involved in testing and training because reference and instruction services will be heavily affected by the installation of a resolver. Customer support by the vendor is also important, and rapid response to users’ needs is particularly critical. He also noted that one does not need to obtain a link resolver and the knowledge database form the same vendor.

Kasproski mentioned the following problems that one might experience during the trial stage:
• Not all holdings may be integrated into the knowledge base, and not all features of the system may be available for testing.
• The level of customer support might not be as high for a trail customer as it would be for a fully paid customer.
• The product could evolve during the trial phase which might cause delays in implementation.

After implementation, the following problems and issues may arise:
• Index problems such as source-target mismatches and gaps in database coverage can affect linking.
• New journal collections may become available, but they may not be added quickly to the knowledge base, so linking is impossible.
• ILL links must be capable of handling all types of content (journals, books, chapters, proceedings, etc.), not only for a single database, but across all platforms.

The Claremont Colleges group is using ArticleLinker by SerialsSolutions, which is a remotely hosted solution. (Budgetary limitations required Claremont to adopt this type of platform.) They chose SerialsSolutions as their vendor because they already had a huge knowledge base that had been built on that platform. One distinguishing feature that they added to the standard system was to create this “huge, obnoxious, green button” for users to click when they wanted to retrieve the full text.

By making the button prominent, they hoped that usage would be stimulated, and it was.

Claremont found that installing a link resolver consumed more staff time than originally planned, and testing and evaluating each knowledge base source was very labor-intensive. Their wish list includes:
• Links to digital dissertations and other non-article content
• Selectable aggregator content (i.e. ability to choose the journals from each collection that are important to them)
• Ability to harvest statistics gathered by the vendors and ensure that they are COUNTER compliant
• Ability to add in-house content to ArticleLinker.

Trainor feels that the future of the OpenURL lies in the ability to link to other types of information, such instant message blogs, MySpace content, mashups, etc.

The University of Louisville uses ExLibris’s SFX link resolver and is happy with it. When the link resolver was installed, the libraries discovered the power of having all electronic journal data in a single place. They would like to see titles that have no SFX targets added to the knowledge base (such as government documents, new titles, and titles lacking ISSNs), as well as e-books.

Don Hawkins
Columnist, Information Today

Comments are closed.