SUBSCRIBE NOW!

Vol. 11 No. 2 — February 2003

By no means does this end the story,however. Linking options and alternatives morphed again into context-sensitive linking or reference linking. No longer is it sufficient to offer a one-to-one link from a bibliographic record or a reference in a bibliography to the corresponding full text housed either internally or externally. Depending on the affiliation of the user clicking on such a one-to-one link, he or she may or may not have rights of access. For example, a link in an EBSCOhost database may be able to take a user from the EBSCOhost database to a publisher's Web site, but unless that user has an institutional or personal subscription, he or she may be denied access. Hence, librarians encounter the "appropriate copy" problem, the challenge of providing a legitimate copy of any given material to users, whether digital or print.

In addition to reaching full text through reference linking, librarians want users to be aware of local print holdings, interlibrary loan, and document delivery options. How can libraries and information centers offer one link, from any source (database, OPAC, e-journal subscription at publisher's Web site), to any appropriate other source (full text, print holdings, other e-journal subscriptions, ILL, document delivery)? Libraries and information centers can offer such dynamic linking through context-sensitive linking by the rather elegant inclusion of a linking technology known as the OpenURL.

The OpenURL framework was developed by Herbert Van de Sompel and a team at Ghent University and by Oren Beit-Arie at Ex Libris. In a nutshell, the OpenURL framework standardizes the transportation of metadata as a syntax for transporting metadata and identifiers. Currently, the standard is undergoing review through a standards committee, the National Information Standards Organization (NISO) Committee AX [http://library.caltech.edu/openurl/]. One of the remarkable aspects of the OpenURL technology is its ability to take "into account the user and the context in which the user is attempting to gain access" to a particular resource.¹[For a fuller explanation of the OpenURL written in accessible language, see Priscilla Caplan's "A Lesson in Linking."² For a technical discussion of the OpenURL, see Mark Needleman's "The OpenURL: An Emerging Standard for Linking."]

Since its development, several services have evolved which use the OpenURL framework to offer seamless, context-sensitive linking. Overall, these services have evolved in an academic environment, attracting academic libraries as their initial customers. However, with the dynamic flexibility of the OpenURL framework, the sky is the limit for all types of information centers. This article will profile four services: Ex Libris' SFX, Endeavor's LinkFinderPlus, Openly Informatic's 1Cate, and the University of North Carolina at Greensboro's Journal Finder. Using the OpenURL framework, all four services introduce a rather simple feature into the world of traditional linking: a middleman. Instead of a one-to-one traditional link from one source (e.g., abstracting and indexing service) to only one target (e.g., full text housed at an aggregator or publisher's Web site), the link now travels through a linking service, which takes the user to one screen containing any number of appropriate copy options.

Ex Libris and SFX

Ex Libris, and more specifically the product SFX [http://www.sfxit.com], pioneered this linking service game, first throwing its hat into the ring about 5 years ago. According to Jenny Walker, director of Sales and Marketing, Information Services Division, "Ex Libris and SilverPlatter were technology partners in the initial SFX research work undertaken at Ghent University in Belgium by Herbert Van de Sompel and his team." In February 2000, Ex Libris acquired the full rights to SFX (short for special effects, named by Van de Sompel) from Ghent University. Walker states that in early 2000 at approximately the same time, Oren Beit-Arie at Ex Libris, together with Van de Sompel, "further developed the concept of the SFX URL and created OpenURL." Both Beit-Arie and Van de Sompel have written several articles about the OpenURL and open linking in general for libraries³.

The SFX linking service allows a library or information center to define its local resources. This sort of localized control translates into context-sensitive reference linking for the user. In order to implement SFX, an information center can purchase a server on which the linking server application or linking resolver runs. Such a server could be used by one library or housed centrally for a consortium. SFX works either as a stand-alone product or with an integrated library system (ILS), such as an ILS sold by Ex Libris or other companies. If an information center chooses not to purchase its own server, it can pay Ex Libris to host the linking resolver on a remote server. Most libraries and information centers, by far, choose the former option to purchase their own servers and host the linking resolver locally.

The localized control for SFX resides in the KnowledgeBase. TheKnowledgeBase is generally a global linking "database containing rules for OpenURL linking, including which journals are available from each publisher/aggregator, coverage dates, and the required linking syntax."⁴ Libraries then upload local subscription and holdings data into the KnowledgeBase (manually or batch), thus customizing the KnowledgeBase to reflect the range of the local collection: digital and print. If a library already uses a service such as SerialsSolutions, then the information provided by SerialsSolutions can be used for a batch upload into the KnowledgeBase. Beyond the local collection, the KnowledgeBase can also be modified to include "extended" services, such as links to interlibrary loan, local FAQs, or preferred document delivery services.

The addition of the link server or link resolver, the middleman, allows libraries and information centers to take users accessing their system from any source — an abstract database search, the OPAC, or others — to appropriate targets — full text, abstract databases, etc. The OpenURL framework supports this dynamic linking. The OpenURL contains the standardized metadata about the material to be retrieved, as well as the affiliation of the user. Harry E. Samuels, Digital Project coordinator for Endeavor Information Systems, provides us with an example of an OpenURL using LinkFinderPlus:
http://LinkFinderPlus.library.edu?genre=article&issn=
12345678&volume=99&issue=1&date=20020101&spage=27⁵. An OpenURL using SFX might look like the following: http://sfx.aaa.edu/menu?genre=article&issn=12345678&volume=
14&issue=9&spage=1&date=1998&aulast=Smith&aufirst=Paul.

The University of Tennessee, Knoxville (UTK) Libraries implemented SFX in 2001 and released it to users in August 2001. Maribeth Manoff, Systems librarian at UTK, notes, "We have run into glitches here and there, but overall it's been a great service." When asked why UTK chose SFX over other possible linking services, Manoff states, "From the research we did, SFX was the most well-known service and it was the one that was ready to go." Manoff went on to comment, "We considered LinkFinderPlus," but at the time when UTK was considering the purchase of a linking service, LinkFinderPlus had not yet been fully released.

SFX has been involved in several key partnerships, particularly with CrossRef from the Publishers International Linking Association (PILA). The OpenURL framework complements CrossRef and makes SFX compatible with CrossRef. CrossRef uses Digital Object Identifiers, or DOIs, and DOIs work as metadata within the OpenURL framework. SFX provides an example of an OpenURL containing a DOI: http://sfxserver.uni.edu/sfxmenu?
id=doi:123/345678&id=pmid:202123
[http://www.sfxit.com/openurl/openurl.html].

In addition to current projects and enhancements, Walker described the following new development:

A recent development includes the ability for two SFX servers — or two SFX "instances" on a shared server — to communicate server-to-server through use of an API. This is being used in consortial environments where shared or core "just-in-time" basis with the individual member institutions. This work has been undertaken initially in conjunction with the California Digital Library, which hosts an SFX server for all CDL shared resources, and one of their member institutions, University of California Davis (UCD), who hosts their own SFX server and configures this only for resources and services that are unique to UCD.

Walker mentioned several other new partnerships and developments for SFX, including early trials with the new OpenURL v. 1.0, once NISO releases it (probably in early 2003) and further SFX development with the Internet2 Shibboleth group related to attribute-based authentication. In terms of cost, SFX pricing is based on a full-time enrollment (FTE) count for academic sites and a population served or registered patron count for public libraries. In general, Ex Libris charges a one-time license fee plus an annual maintenance charge.

Endeavor Information Systems and LinkFinderPlus

Endeavor Information Systems sold its first Voyager system, an integrated library management system, to Michigan Technological University in 1994. Since that time Endeavor has expanded its services to include a variety of products available for separate purchase or as add-ons to the Voyager system, including an interlibrary loan module, a citation server, and more. The latest service to join the Endeavor family of products is LinkFinderPlus [http://www.endinfosys.com/prods/linkfinderplus.htm], Endeavor's link server application. According to Penny Emke, Marketing Communication manager of Endeavor Information Systems, Endeavor recognized that "users were frustrated with the multi-step process of looking up a citation in an A&I database, determining if the journal is in the OPAC or going to a separate Web page list of e-journals to discover if the library has access to the journal." LinkFinderPlus was designed to meet the changing needs of library users. "LinkFinderPlus bridges the gap between citations and the online full-text of the article, providing maximum usage of the library's electronic journal subscriptions," states Emke.

Released in June 2002, LinkFinderPlus is available as a stand-alone product or as part of EnCompass, Endeavor's digital library system. As of November 2002, when Endeavor introduced LinkFinderPlus 3.0, Endeavor had 41 EnCompass installations and 52 LinkFinderPlus installations worldwide [http://www.endinfosys.com/news/lfp3fin.htm]. Typically, libraries or information centers using LinkFinderPlus purchase a server and house the server themselves. For these customers Endeavor usually installs the linkingserver software remotely. However, a customer may send its server to Endeavor to have the LinkFinderPlus software installed. Upon completion of the installation, Endeavor returns the server to the customer. Endeavor also offers a remote hosted option, wherein the company will house the server⁶.

LinkFinderPlus comes with a pre-configured knowledge database of over 13,000 journals. To indicate which particular full-text resources it subscribes to, a library or information center need only check or un-check boxes within the knowledge database. Rachel Frick, head of Bibliographic Access Services at the University of Richmond, notes, "Once I walked through the steps of setting up the knowledge base I was able to just roll with it and activate all of our available resources. Very easy to learn." Using the administrative client, LinkFinderPlus customers can further customize the services provided to library users. The library may rank the resources that it would like to provide to its users, e.g., to present the most appropriate copy first, and add extended services beyond electronic full text, e.g., links to interlibrary loan services, search engines, and online booksellers.

To enhance the services provided to libraries using both SerialsSolutions and LinkFinderPlus, Endeavor recently established a partnership with SerialsSolutions "that will allow mutual clients to easily transfer and integrate data between their respective content management systems." According to Emke, "Libraries can use information from services like SerialsSolutions to activate targets from within the LinkFinderPlus Knowledge Base...which saves the library time in implementing and maintaining their linking system."

Like SFX, LinkFinderPlus relies upon and complements industry standards and initiatives, such as the OpenURL and CrossRef. Adds Emke, "By employing the OpenURL protocol, LinkFinderPlus offers intuitive linking from citations to full-text, additional resources, and 'more like this' extended services." When asked about the current pricing structure for LinkFinderPlus,Emke states that "agreements with our customers restrict Endeavor from releasing pricing information." However, Emke notes that LinkFinderPlus is a competitively priced system.

Openly Informatics and 1Cate

Openly Informatics has also developed a linking service, 1Cate [http://www.openly.com/1cate/]. Eric Hellman, president of Openly Informatics, founded the company in the late 1990s. Formerly a research scientist at Bell Labs, Hellman became involved in linking technology while building an automated e-journal publishing system for the MRS Internet Journal of Nitride Semiconductor Research. Openly Informatics first entered the linking service arena through involvement in jake (the jointly administered knowledge environment), a free software initiative maintaining information about databases and aggregated collections. 1Cate, which stands for 1-click-access-to-everything, grew out of Openly Jake, a framework for customizing data from jake. However, Openly has moved away from jake and expanded its link server products to interact with other data services⁷.

Openly hosts the 1Cate linking server or linking resolver, also called Hosted 1Cate, remotely. The subscribing institution provides Openly with information about the library's configurations and data sources, as well as Web page design. Using its link server, Openly creates a link server Web site based upon the institution's specifications. Under this arrangement, the library does not purchase any hardware or software.

Hosted 1Cate offers a number of features, one of them the automatic creation of Alphabetical Journal Lists. Using the list of data sources that the library or information center provides, 1Cate will automatically provide alphabetical title lists with a word search capability. The Hosted 1Cate package also includes the JournalSeek Linking Database. This database, created through a partnership between Openly and Genamics.com, provides journal descriptions and linking information for over 9,000 e-journals. Another useful module, the Link.Openly Server, "uses open XML standards to make 'deep' links to e-journals at the volume, issue, and article level." According to the Openly Informatics Web site, 1Cate offers UltraCustomization, making 1Cate look and feel like it is part of the library's Web site. If using the standard link server does not meet a library's needs, Openly Informatics offers Custom 1Cate, a service in which Openly will customize a link server based upon a library's unique requirements.

Instead of paying a purchase price and a monthly maintenance fee, 1Cate is a leased service. Marketing 1Cate primarily to academic libraries, Openly Informatics bases its pricing upon the full-time enrollment (FTE) of the academic institutions, as well as the number of products and features the library wishes to include. There are three pricing tiers: basic, middle, and research. According to Hellman, Openly Informatics shortly expects to announce sales of 1Cate through an independent reseller.

While Openly Informatics does not offer an ILS in conjunction with 1Cate and is not affiliated with any particular ILS vendor, it has several ILS vendors as customers. According to Hellman, Openly Informatics has concentrated its efforts on developing the technology behind its client software, and ILS vendors have purchased these pieces of software to use inside their products. For example, Endeavor Information Systems entered into partnership with Openly to provide extended access to resources through its own link server system, LinkFinderPlus. Endeavor integrated the JournalSeek database, as well as the Link.Openly service, into LinkFinderPlus.

University of North Carolina at Greensboro and Journal Finder

Unlike the previous three commercial linking services, an academic institution, the Walter Clinton Jackson Library at the University of North Carolina at Greensboro (UNCG), created Journal Finder and released it to the public in August 2001. As described by its creators, "Journal Finder is a locally developed software product that lets patrons search for a journal title and receive several options for that title."⁸ Much like SFX, LinkFinderPlus, and 1Cate, Journal Finder works as a behind-the-scenes linking service application that allows the library to offer one-stop shopping for all its digital and print collections, as well as document delivery, ILL, and remote library catalog options.

When asked what prompted UNCG to create its own linking service, Tim Bucknall, Assistant director of Information Technologies and Electronic Resources, replied, "At the time we built Journal Finder, SFX was in beta and was not commercially available. LinkFinderPlus wasn't even in development, as far as I know.... We had already been doing the content portion (i.e., the e-journal holdings) for some time. ... We also thought SFX was quite expensive for what it did."

Journal Finder functions both as a linking resolver and an A through Z listing of electronic serials (very similar to a product like SerialsSolutions). The linking resolver works based on the OpenURL framework, and the A through Z listing contains dates of coverage for serials from different providers, thus allowing users to access full text. Bucknall notes, "The main disadvantage to Journal Finder is that we make no attempt to provide global content for as many full-text e-resources as possible. Our content is limited to the 100+ full text products to which UNCG has access."

Journal Finder also has some unique features, such as the ability to "correct" user typing mistakes. Bucknall explains:

Journal Finder doesn't just search on the journal title string that the user enters. We run it through a filter that removes "the," "and," "&," and most punctuation, and chops off journal subtitles. The "improved" search string is then searched against a field that contains titles, title abbreviations, and title variants. If the user makes a typing mistake and we find similar (but not exact) matches in our database, then the user is alerted to that fact and redirected to those similar titles.

Users can access Journal Finder and its linking capabilities from other databases, the UNCG OPAC, or can search it directly by visiting http://journalfinder.uncg.edu/uncg. According to Bucknall, future plans for Journal Finder include expanding and refining OpenURL linking capability as well as adding corporate authors and more periodical title abbreviations along with added titles.

Because Journal Finder was developed specifically for UNCG and the university uses the DRA system, it is compatible with the DRA Web 2 system. However, Bucknall states that Journal Finder should work fine with Ex Libris, Endeavor, or most anything else. The caveat is that there are some different options for integrating Journal Finder with different ILSs and different technical services workflows, and some options consume more time than others.

UNCG does want to collaborate on its product with other institutions. Currently, Journal Finder is in use at six DRA sites and four Innovative Interfaces Incorporated sites. In terms of pricing for collaboration, there are two options: 1) consulting, at $5,000 plus expenses, and 2) hosting, at $1 per FTE per year, with a minimum of $1,000 per year.

We've Come a Long Way

The four services profiled above are not the only linking services in town. According to Albert Simmonds, Business manager, Registry & Metadata Services at OCLC, OCLC has also developed its own OpenURL-compliant linking service, OCLC Cooperative Rights & Resolution Service (R & R). OCLC's R & R is a cooperatively developed central repository for rights information. A library registers holdings for its print and electronic serials, as well as its aggregator databases. The R & R service compares the rights metadata for those sources to user requests and delivers the appropriate copies to users. Pending approval by the OCLC management team, the R & R service is tentatively scheduled for release in April 2003.

Regardless of vendor, the main goal of all linking services is never to leave the user at a dead end, to provide users with access to their appropriate copies. If digital or print subscriptions are not an option, then the service should offer the user ILL or document delivery, or even access via another library's catalog. When we consider the sophistication of these systems and their capability to link seamlessly among differing resources, we also need to think about how far we've come in a relatively short period of time. As a gauge, the number of publications listed in Fulltext Sources Online (FSO) has grown from approximately 4,400 in 1993, to 13,084 in July 2000, and as of July 2002, the main section of FSO contains entries for 17,467 publications. In 1993, less than 10 years ago, services such as SFX were not even a gleam in their creators' eyes. We've certainly come a long way, and we have linking advances and services not yet charted.

References Brandsma, Terry W., Elizabeth R. Bernhardt, and Dana M. Sally, "Journal Finder: A Solution for Comprehensive and Unmediated Access to Journal Articles," Serials Review, vol.28, no.1, 2002, pp. 13-20. Caplan, Priscilla, "A Lesson in Linking," Library Journal NetConnect: Supplement to Library Journal and School Library Journal, vol.126, no. 17, Fall 2001, pp. 16-18. Collins, Maria and Christine L. Ferguson, "Context-Sensitive Linking: It's a Small World After All," Serials Review, vol. 28, no. 4, 2002, forthcoming. Needleman, Mark, "The OpenURL: An Emerging Standard for Linking," Serials Review, vol. 28, no. 1, 2002, pp. 76+. Samuels, Harry E., "What Is an OpenURL?," Available at http://www.endinfosys.com/pdf/openurl4_02.pdf. Van de Sompel, Herbert and Oren Beit-Arie, "Generalizing the OpenURL Framework Beyond References to Scholarly Works," D-Lib Magazine, vol. 7, no. 7/8, July/August 2001; available at http://www.dlib.org/dlib/july01/ vandesompel/07vandesompel.html. Van de Sompel, Herbert and Oren Beit-Arie, "Open Linking in the Scholarly Information Environment Using the OpenURL Framework," D-Lib Magazine, vol. 7, no. 3, March 2001; available at http://www.dlib.org/dlib/march01/ vandesompel/03vandesompel.html.

Footnotes

---

¹ Mark Needleman, "The OpenURL: An Emerging Standard for Linking," Serials Review, vol. 28, no. 1, 2002, pp. 76+.

² Priscilla Caplan, "A Lesson in Linking," Library Journal NetConnect: Supplement to Library Journal and School Library Journal, vol.126, no. 17, Fall 2001, pp. 16-18.

³ Herbert Van de Sompel and Oren Beit-Arie, "Generalizing the OpenURL Framework Beyond References to Scholarly Works," D-Lib Magazine, vol. 7, no. 7/8, July/Aug. 2001. Available at
http://www.dlib.org/dlib/july01/
vandesompel/07vandesompel.html; Herbert Van de Sompel and Oren Beit-Arie, "Open Linking in the Scholarly Information Environment Using the OpenURL Framework," D-Lib Magazine, vol. 7, no. 3, March 2001. Available at
http://www.dlib.org/dlib/march01/
vandesompel/03vandesompel.html.

⁴ Maria Collins and Christine L. Ferguson, "Context-Sensitive Linking: It's a Small World After All," Serials Review, vol. 28, no. 4 , 2002, forthcoming.

⁵ Harry E. Samuels, "What Is an OpenURL?" Available at
http://www.endinfosys.com/pdf/openurl4_02.pdf.

⁶ Collins and Ferguson, Serials Review.

⁷ Collins and Ferguson, Serials Review.

⁸ Terry W. Brandsma, Elizabeth R. Bernhardt, and Dana M. Sally, "Journal Finder: A Solution for Comprehensive and Unmediated Access to Journal Articles," Serials Review vol.28, no. 1, 2002, pp. 14+.