Which Route Do I Take?  A Viewpoint on Locally-Developed vs. Commercially Available Journal Management Solutions

Access to Journal Literature:  A Brief Overview

Ready access to journal literature has long been a critical component of successful library services.  In the print journal environment, this used to be a somewhat routine or stable process, if not particularly user-friendly.  Although this is a greatly simplified description of the “old” process, generally, access to print journal literature for the user involved two main steps.  To find articles, users first consulted appropriate print or online indexes to obtain citations to journal articles.  Next, they searched the online catalog to determine if their library held the issues of the journal referenced in those citations in order to get to the desired articles.  However, in these days of dramatically increasing availability to libraries of journals in online or electronic form in full text, the simple paradigm I have described is increasingly irrelevant for users.  Our users now expect their literature search to be a one-stop shopping experience, with the ability to search an online database for relevant citations along with the ability to quickly and easily navigate directly to the full text of the articles from their citations.

Most libraries catalog their print journals to provide subject and descriptive access for their users.  Many do the same for e-journals to which they provide access.  In addition, many libraries have been developing web-based A to Z lists of the journals to which they provide access for several years now, particularly for e-journals.  Generally speaking, those who use these lists can navigate directly from a title-level link to the full text of the articles that are available online for a particular journal, whether from an aggregator database, publisher’s Website, or on the Web.  This service, although popular with users, presents several challenges for libraries.  For one thing, aggregator database content is constantly changing, making it difficult for libraries to keep their web-based A to Z lists up-to-date.  Also, it is quite common for online content for a particular journal to be available from several different sources, each with varying dates of coverage or possibly even different types of content (e.g. one vendor package may offer the journal articles in ASCII text, while another may offer it in HTML and/or PDF).  Keeping track of where a particular journal is available to authenticated library users, and how to succinctly document for users the range of different places in which they can access them, represents another big challenge for libraries.

Solutions to Journal Access Issues

Libraries have tried various approaches to improve access to print and online journals, both from the users’ perspective in terms of ease of use and accuracy of information, as well as from the maintenance end of things.  Locally-developed solutions seemed to be the first step, with an emphasis on database-driven applications such as the Electronic Resources Database developed at Griffith University, Queensland, Australia; as well as HERMES, developed at Johns Hopkins University.^1,2  Another noteworthy, locally-developed system is Vera, developed at MIT.³  Tim Jewell of the University of Washington and Adam Chandler of Cornell University maintain an excellent Website describing these and several other local initiatives, called “A Web Hub for Developing Administrative Metadata for Electronic Resource Management,” which is part of a larger effort funded by the Digital Library Federation to identify and standardize on administrative metadata for e-resources.⁴

Commercial vendors have not been idle during this period of activity with regard to local solutions.  They have been busy creating and marketing their own journal or e-resource management services in response to customer demand.  Integrated library systems vendors such as Endeavor, Ex Libris, SIRSI, and Innovative Interfaces, along with subscription vendors such as EBSCO and vendors focused exclusively on this particular niche, such as TDNet and Serials Solutions, have all recently introduced products to the library marketplace whose aim is to assist librarians in better managing their journal collections.⁵  Such applications offer many significant advantages for libraries, especially in terms of convenience and added value (e.g. by integrating OpenURL services for citation linking), in much the same way that subscription vendors provide a valuable service as the intermediary between libraries and journal publishers, saving libraries the cost of managing journal subscriptions individually.  Although concrete numbers are not readily available at this point in time, anecdotal evidence suggests that this new kind of vendor-supplied service has quickly become very popular with libraries.  A comprehensive survey of the library marketplace to document the penetration of commercially-provided journal management solutions would make a good topic for a separate article.

Whither the Locally-Developed Solution?

The meteoric rise of commercial solutions to managing journals in the last few years raises the obvious question of whether or not a local solution is desirable for libraries to pursue any more.  Many libraries which once had locally-developed solutions have abandoned them in favor of vendor-supplied services.  Locally-developed journal management solutions can be resource-intensive to create and maintain, and many libraries, even large academic libraries, may lack the staff and/or technical resources to adequately support such a service.  For instance, keeping up-to-date in terms of journal coverage, especially in aggregator databases, is notoriously difficult.  Local solutions may also lack extensibility to adequately support the dramatically increasing number of journals that are available to a particular library, or they may not be readily adaptable to related services such as OpenURL linking.

On the other hand, are commercial solutions really a better alternative?  Is there really any need to debate the issue, or are locally-developed journal management solutions a lost cause in the same way that locally-developed integrated library systems have gone out of fashion in favor of vendor-supplied systems?  Some negative aspects to vendor-supplied solutions include expense and the lack of local control.  Although costs vary widely from vendor to vendor and also in relation to many other factors such as the number of journals a library wishes to manage, vendor solutions are generally subscription-based, meaning that libraries can expect to pay a significant amount of money per year to subscribe to a vendor-supplied solution.  And although they are making great strides to enable more of it, capabilities for local customization (e.g. in how the journal information is presented to local users) in vendor-supplied journal management systems can be rather limited in comparison to a local solution.

Clearly, for libraries who desire a tool to help them improve access to journal literature for their users, especially for e-journals, there are many variables to consider when deciding which route to take, and those that have been mentioned thus far just scratch the surface.  In spite of the attractiveness of vendor-based solutions, there are many good reasons for taking the route of developing and maintaining a local system, particularly for small college and university libraries, as well as some public libraries.  This is what the author’s library has chosen to do.  The second half of this article describes the path the author’s library has taken as a case study of possibilities for locally-developed journal management systems.

Spotlight on One Library’s Local Solution

Two years ago, Taylor University in Upland, IN, decided to implement a locally-developed system to provide better access to its growing array of print and online journals.  Taylor is a small, evangelical Christian institution with approximately 2,400 undergraduates and a small but growing number of graduate programs.  The initial system was a database-driven application using ASP as the query language and a database stored in Microsoft Access mounted as a file on our institution’s external Web server.  An undergraduate computer science student, who was looking for real-world experience in database design and Web application development, was hired to develop the system.  The initial database was propagated using separate vendor title lists from each vendor to which Taylor subscribed, combined with a title list of print and microform periodicals maintained on an Excel spreadsheet.  The user search interface, known as the Taylor Periodical List, initially allowed the user to browse the periodical list by the first letter of the title but later on, the ability to do title keyword searches as well as restrict by format (e.g. print, online, microform) was added.  Once the user found the title he or she was looking for, if the periodical was available online, a hypertext link with the name of the database as link text was presented on which the user could click to go directly to the articles available in that particular database.

Figure 1. Taylor Periodical List Search Results

All links to licensed journal content were directed through EZProxy to ensure authenticated access to on- or off-campus users.  In addition, the user was given full-text dates that indicated coverage for e-journals in the database.  In the case of print or microform periodicals, the user was presented with a hypertext link to the library’s online catalog so that current holdings information could be ascertained by means of a preconstructed ISSN search.

The Taylor Periodical List quickly became a “mission critical” application, offering access to around 4,000 journals (about half were print or microform titles, the other half were online, mostly from aggregator databases such as EBSCO Academic Search Premier).  Website statistics showed that the Taylor Periodical List was heavily used, second only to the online catalog in popularity.  In spite of this, the system had several limitations, both from the user perspective and from the maintenance end of things.  For example, the amount of manual effort required to combine several disparate vendor title lists on a regular basis was extensive and expensive, even though it was mostly accomplished with part-time student help.  Standardizing different ways of representing a title (e.g. something as simple as vendor X giving a title as “Journal of…” while vendor Y gave the same journal the title of “The journal of…” which resulted in duplicate entries in the list for the same journal) was a big headache, as was resolving the different ways in which title coverage dates were handled from vendor to vendor.  Also, subject access was not provided as part of the interface.

Planning began in early 2003 for a revised, improved periodical management system.  Enhancement objectives for the new system, which was named Taylor Periodical Administration System (or TPAS for short), included the following:

At around the same time as this plan developed, Taylor first began using Ex Libris’s OpenURL-based, context-sensitive linking solution, known as SFX, which became available as part of a consortial purchase (Taylor is part of a consortium known as PALNI (Private Academic Library Network of Indiana)).  An additional goal of the TPAS project was to integrate SFX and TPAS where possible, particularly in regard to utilizing SFX's regularly updated Knowledge Base from which to derive up-to-date title list information for periodical databases to which Taylor subscribes.  This step would drastically reduce the amount of maintenance required to keep TPAS up-to-date.  Furthermore, it would help to provide accurate dates for journal availability, along with embargo information.

The same undergraduate computer science student who worked on the initial implementation developed TPAS during the summer of 2003, and it was first unveiled for users in the form of a new Taylor Periodical List interface in October 2003.  The system now runs on a server running Microsoft SQLServer 2000 using PHP as the query language.  The user interface is simplified and easier to use.

Figure 2.  Initial Search Screen of Taylor Periodical List

The number of periodicals maintained in the system has jumped to about 11,000 unique titles from thirty seven different sources or databases (the number of print and microform titles included in this number is roughly 2,000).  Several administrative data elements are now maintained in TPAS including vendor information, consortium information (if we get a group of journals through a consortial subscription, e.g.), database-specific information, departmental affiliations, subscription information, and usage statistics as well as other kinds of statistical data.

Figure 3.  TPAS Database Update Screen

Data in the periodical database is updated on a monthly basis by means of data imported from the SFX Knowledge Base.  Among other elements of this updating process, journal titles that have been dropped from a particular database are removed automatically, and information about any new titles that have been added, or new date coverage information, is added or refreshed automatically.

A particularly useful element from the user perspective is the addition of subject searches to the user interface.  After considering various methods for providing this kind of access, Taylor chose to use a simplified subject thesaurus based on degree majors offered locally, augmented by some additional subjects such as “General” or “Newspapers.”  This choice was inspired by a similar approach taken by the University of Victoria (British Columbia, Canada) Libraries, although their method of subject assignment differs significantly from Taylor’s.⁶  This simplified subject approach was felt to be beneficial for several reasons, not least of which was to enable closer mapping of available journal literature to the curriculum offered at Taylor.  For example, a link to a comprehensive list of all math-related journals could easily be generated and used in class links in Taylor’s course management system (Blackboard).  See table 1 for the complete subject list.

Table 1. List of Subjects Used By Taylor University

In order to populate subjects in the database, an extensive list of title keywords for each subject is defined and that list is then used to automate the subject assignment process.  This approach to providing subject access is still highly experimental and simplistic.  Approximately half of all periodical titles available through the Taylor Periodical List now have at least one assigned subject.

As a representative system, TPAS has lots of room for improvement, and more enhancements are being planned.  During the next year, developmental efforts will focus on wrapping the various system components into an open source package, readily available for anyone to download and install from SourceForge.net, complete with detailed instructions.  Contact Steve Oberg at stoberg@tayloru.edu if you are interested in finding out more about TPAS.

Conclusion

It takes a great deal of effort to effectively manage journals, particularly with the rapid growth in access to e-journals in many libraries.  E-journals are no longer a sideline but instead are mainstream in terms of their role in library collections, especially since the rise in popularity of aggregator databases.  The combination of managing various forms of e-journals along with existing traditional collections can be quite daunting.  Journal management systems provide a new opportunity to simplify that management process as well as improve access for users.  Recent trends seem to favor vendor-supplied solutions, but do not count out locally-developed systems just yet, particularly as they become more full-featured, standardized, and freely available as open source.

Notes