Electronic Journal Forum SEESAU: University of Georgia's Electronic Journal Verification System Maria Collins and William T. Murray, Contributors Maria Collins, Column Editor Available online 5 April 2009
In support of its increasing electronic journal collection, William T. Murray from the University of Georgia Libraries has developed an access verification tool called SEESAU (Serial Experimental Electronic Subscription Access Utilities). This Web interface has utilized UGA's integrated library system and link resolver to coordinate approximately 37,000 access checks since January 2007. A problem resolution system is included as part of SEESAU and functions similarly to a ticketing system. Through use of SEESAU, library staff have not only discovered access problems proactively, but they have also discovered Web site performance issues, title maintenance concerns and corrections that need to be made to their knowledgebase. Serials Review 2009; 35:80–87. © 2009 Published by Elsevier Inc.
Access problems can be an everyday reality for libraries with electronic journal collections. Due to the large number of titles to which many libraries subscribe, a proactive approach to maintaining access to electronic journals may seem quite daunting. Addressing access problems as they are discovered, or fix upon failure, is a common strategy employed by many libraries with limited resources and personnel. Unfortunately, all the staff time saved through the fix-upon-failure strategy may not be worth the patron ire and frustration these access problems can engender. Library staff from the University of Georgia (UGA) Libraries have made a deliberate decision to abandon the fixupon-failure concept and have experimented with a locally developed system called SEESAU to verify access to their electronic journal titles. William T. Murray, the developer of SEESAU, stated that UGA first began discussions of building an access verification tool in 2003, when the library began a large-scale project to convert print subscriptions to online. As print subscriptions were replaced with online, consistent access problems began to occur, calling into question the library's ability to provide reliable and timely access to journal content. Furthermore, given the ever-increasing budget allocation required to maintain a serials collection of UGA's size, issues of inventory control, or rather making sure the library was getting what it paid for, provided further impetus to develop a systematic approach to proactively ensure online availability. By the end of 2006, enough access problems had occurred to provide momentum for the project and development began in earnest. A team of five people assisted Murray with SEESAU's design, and the
tool was successfully launched on January 8, 2007. In fact, the first access check occurred at 11:56 a.m. and access to the Journal of Differential Equations was confirmed. Since January 2007, the SEESAU system has performed 37,000 access checks for approximately 4,000 core journal titles out of the 50,000 to which the library has access. These 4,000 titles are ones that UGA wants to carefully track and do not include aggregator titles, free titles or ‘extras’ provided to the library through certain packages.
How Does SEESAU Work? SEESAU is a Web-based interface (http://www.hindcite.com/ SEESAU1.htm) that uses data harvested from UGA's Voyager ILS (integrated library system) for title identification and scheduling and then uses the library's link resolver SFX to run the access check. Data are extracted from Voyager using Microsoft Access and then run through a series of data processing programs written in Perl before being uploaded to the SEESAU server where CGI (Common Gateway Interface) scripts take over. SEESAU constructs an HTML (Hypertext Markup Language) link for each title record loaded from the ILS data that passes the user back through the link resolver to perform the access verification check. Murray has set this process to run automatically once a week on Friday night, and titles are uploaded to the SEESAU server every Monday morning. At this point, no routine maintenance is required for these reports to run. Data preparation time to upload titles to SEESAU each Monday morning takes approximately five minutes. The SEESAU server itself takes about two hours to process and prepare these weekly loads. Figure 1 further illustrates the workflow between SEESAU, Voyager and SFX. The data matchpoints between SEESAU and Voyager are the bibliographic record ID number, the purchase order number and
Collins is Assistant Head of Acquisitions, NCSU Libraries, North Carolina State University, Raleigh, NC 27695, USA; e-mail:
[email protected]. Murray is Serials Public Services Section Head, University of Georgia Libraries, Athens, GA 30602, USA; e-mail:
[email protected]. 0098-7913/$ – see front matter © 2009 Published by Elsevier Inc. doi:10.1016/j.serrev.2009.02.003
80
Collins / Serials Review 35 (2009) 80–87
Figure 1. Flowchart of SEESAU data model.
the check-in record component name. Other codes pulled into SEESAU, such as the PUBBY code and PreferredTarget, are used as validation points. The PUBBY code, which is created by UGA's serials staff, represents the publisher name; the PreferredTarget is the publicly seen name of the SFX target. Both of these codes are manually recorded in Voyager on the line item in different places as shown in Fig. 2. The PUBBY code provides an additional benefit from an auditing point of view, as it provides a traceable path from the access point and appropriate copy of a resource to the payment information. This connection is an important and unique feature of the SEESAU system, distinguishing the version or collection affiliation of a title purchased. Often, in integrated library systems, acquisitions-specific distinctions between versions of a given title purchased are usually only artificially created through separate order records and notes. With the PUBBY code, UGA has found a way to standardize a description of the appropriate copy for a given title. In order to systematically schedule titles for access verification, SEESAU relies on the Voyager serial record, which includes a pattern or prediction record traditionally used to record the publication cycle and create check-in records. In this instance, the pattern record is used to stagger the titles that need to be searched and verified on a quarterly basis throughout the year. If an individual wished to verify a title less or more than four times a year, he/she would only need to adjust the pattern according to a desired schedule. To further illustrate how SEESAU works, here is an example of the workflows involved to add a new title to the interface. Like all other new serials
included in the catalog, the title would receive a bibliographic record and purchase order. The order would include any special codes SEESAU might need such as the PUBBY code or PreferredTarget information. A check-in record would be generated and a pattern record created according to the number of desired access checks per year. The expected dates generated from the pattern record would determine when the title would appear in the report messaged in Access and uploaded into SEESAU. Simultaneously, the title would need to be turned on in the appropriate target within SFX. At this point the title will automatically be included in the text file that gets processed in Access according to the schedule set by the pattern. No additional updates and changes are required in SEESAU to pull in the new title.
Working with the SEESAU Title Lists SEESAU distributes the titles across six groups for searching. The first five groups are for specific library staff. The sixth group is a catch-all silo that any staff member can work. On average, each of the five staff members will have seventy-five titles to work each week. On average, the initial title verification process usually takes around five hours, per person, per week. Each link should be followed as far as the staff member can go until full content to article-level access is obtained. As staff work through a year's cycle of titles, they will be assigned the same titles. This familiarity should assist with developing expertise in problem solving.
81
Collins / Serials Review 35 (2009) 80–87
Figure 2. Voyager with PUBBY and preferred target codes.
Figure 3 shows the access check form that the staff see as they begin to work their assigned titles. The number of access points varies, depending on how many ISSNs (International Standard Serial Numbers) are in the MARC 022 field of the bibliographic record in Voyager. There should be at least one title search and at least one ISSN; but often there are multiple ISSNs in the MARC 022 field, creating additional search options in SEESAU. If there are no alternate ISSNs in the 022, the alternate ISSN search option does not display. When the “Go to EJL Search” is clicked, no search is performed automatically; the SFX interface opens to a search form in the electronic journal locator (EJL) for the staff member to use in constructing a title search. This option is used when the “Try it” links fail. “Try it” links may fail if the local title (in UGA's bibliographic record) is different from the publisher's version of the title, or if none of the 022 ISSNs matches the ISSN used by the publisher. By manually searching the EJL, particularly using keyword options, staff can usually find the right title. Non-English language titles with diacritics are problematic for the “Try it” search. This is the reason for a profile option in SEESAU to allow UGA library staff to normalize a title to be in agreement with SFX if such a failure is discovered. If access is verified successfully, the staff member will delete the record from the work list. If an access problem is discovered, the staff member will complete SEESAU's problem form using the Situation Report Utility. This utility, also illustrated in Fig. 3, allows the staff member to describe the
problem according to preset categories or describe the problem as “other.” Notes can be included to provide additional details. At this point, the title is pushed into SEESAU's problem resolution workflow. As each staff member completes a search in SEESAU, results are recorded in Voyager whether successful or not. The staff member will install coded phrases in the check-in issue information field and date the entries. This information allows performance data for a title to be harvested. The most common code used is “access confirmed.”
Problem Resolution in SEESAU Once a title has been designated as a problem using the Situation Utility Report, the associate department head reviews the problem through a separate Problem Workform interface (see Fig. 4. The Main Problem Workform). Within this interface, she can triage the problem in the following ways: she can handle it herself and close out the problem; if the problem is directly related to SFX, she can assign the problem to the SFX liaison, who is responsible for maintaining the SFX knowledgebase; or if the problem cannot be resolved immediately, she can code it as pending and come back to it later. If the problem is sent to the SFX liaison, a separate interface exists for review (see Fig. 5. SFX Liaison Problem Form). If a problem requires more in depth work or coordination with other parties, a Problem Tracking Utility, which functions as a ticketing system, can be used. Note that
82
Collins / Serials Review 35 (2009) 80–87
Figure 3. Access check form.
this is an independent utility separate from the Situation Report Utility, which is the primary means of passing information about a given problem from one unit to another. The Problem Tracking
Utility, however, is used to record work in progress on a particular problem, independent of who the problem has been bounced to through the Situation Report Utility. Currently, the SFX liaison is
Figure 4. The Main Problem Workform.
83
Collins / Serials Review 35 (2009) 80–87
Figure 5. SFX Liaison Problem Form.
responsible for managing the Problem Tracking Utility. The utility is also useful for tracking larger problems that affect entire groups of titles. Within the Problem Tracking Utility (see Fig. 6), each problem is assigned a unique number and one of seven statuses including New, Pending, Resolved, Delayed, Expired Delay, Unresolved, or Bypass. Problems with a “New” status appear at the top of the problem queue. Resolved problems fall in problem number order at the bottom of the list. If the problem requires follow up action at a later date, the problem can be delayed by assigning a date within the problem record for future follow up. Once the delayed status has expired, the problem will again post to the top of the queue with the new problems and be given the status of “Expired delay.” If a particular problem affects numerous titles, such as a sitewide problem with a particular publisher, an individual can create a bypass for that problem. A separate ByPass form is available within SEESAU to manage this information in the system (see Fig. 7). Essentially, the bypass will prevent the same problem from getting reported again and again by telling the system that titles for a particular publisher within a particular target should not be reported in the problem list for review. If the user is aware of a specific date when the access problem will be resolved, a lift date can be set within the ByPass form for that particular bypass to expire. Note that a bypass does not affect the titles coming into SEESAU to be verified. Titles will still come into the SEESAU system as scheduled and searched for other potential access problems. Also, if a staff member learns of an access problem from a
source outside of SEESAU, problems can be created within the Problem Tracking Utility itself.
Types of Problems Discovered Of the 37,000 access checks made over two years, approximately 2 percent were reported as serious problems. To provide greater perspective concerning this failure rate, consider that 2 percent equals one-and-a-half title failures per day. Essentially, this means that every day at least one title is unavailable to patrons. If this same failure rate were applied to all of UGA's titles and not just the 4,000 core titles selected for access verification, this would translate to approximately 3,150 title failures per year. Staff experienced a variety of errors while verifying access including the following:
• • • • • • •
being denied access to content; receiving a message that access is not authorized; sitewide access denial; inability to locate the title at a publisher site; current issues being blocked; all but the current issue being blocked; and issues being loaded incorrectly or delayed.
These problems included both order and link resolution issues. SFX problems identified required additional coordination of targets and thresholds for cleanup. Several of these problems revealed title maintenance issues, such as title changes, frequency changes, publisher changes and ceased titles. Murray believes that
84
Collins / Serials Review 35 (2009) 80–87
Figure 6. Problem Tracking Utility.
these kinds of issues were discovered earlier than usual through the use of the SEESAU system due both to the errors discovered through access checks and notices posted to publisher Web sites. UGA has not yet systematically categorized the problems discovered; therefore, statistics are not yet available on how many of these issues were technical versus bibliographic problems or changes.
Working with a system like SEESAU that harnesses the functionality of two different systems and makes use of data that already exist also teaches staff about how systems and records relate. Understanding how data connects or matches between systems is crucial to the success or failure of a system like SEESAU. This kind of intimacy with the data cannot be underestimated, especially since automation of processes such as access verification is dependent on data consistency. This kind of data intelligence can only serve UGA staff well when and if they implement an electronic resource management system (ERMS).
Benefits of Using the SEESAU System In addition to the ability to discover and fix access problems proactively, SEESAU has given serials staff at UGA a better understanding of the issues involved in electronic journal management. In a sense, this tool has allowed staff to gain a sophisticated knowledge of the collection. The access verification process has revealed to staff performance issues for certain publishers and certain journals. The access check results recorded in the Voyager system provide a historical record of this title activity. This collection intelligence is made possible by the tight link between the title for which access is verified and the acquisition and serials records in Voyager. Essentially, problem reports are linked directly back to the order record, which details information about the publisher, vendor and other bibliographic data. Staff also become very familiar with other performance issues that are not necessarily recorded as a problem in SEESAU. For example, certain publisher sites might always run slow or access to a title may take ten clicks of the mouse. These kinds of details help staff build a certain expertise and comfort level with the library's electronic journal collection.
Future Developments and Related Projects SEESAU is an experimental project; UGA wanted to see if a systematic approach to access verification was possible and if so, how the process could be carried out efficiently. No current enhancements are planned for the basic functionality of the system, although Murray is currently working on a few cosmetic interface changes that will be available by the beginning of 2009. UGA, however, has implemented or is investigating several related projects. Following is a brief summary:
• Archival Access Verification: SEESAU is designed to work with active titles on order but not with titles that are closed out due to a cancellation or cessations; however, many electronic journals with an inactive status of cancelled or ceased, still have valid online holdings. Murray has completed a separate program designed to pull in these inactive titles for access
85
Collins / Serials Review 35 (2009) 80–87
Figure 7. Problem Tracking Utility ByPass form.
verification. See Figure 8 for an example of this new project, which will be launched with the staff by the beginning of 2009. At a future date, this project may be merged with SEESAU; • Connecting SEESAU to usage statistics in SFX: UGA is investigating the potential of reviewing usage data for titles verified in SEESAU to allow for more sophisticated evaluation of use patterns. This information should allow staff to tailor access checks according to the use pattern of a given title. Staff can then increase access checks for some titles and decrease these checks for others; • Claiming project: Similar to the SEESAU project, UGA has created an interface for print claiming. UGA's Claim Wizard has been in operation since 2001 and is used to claim all formats of serials, including microforms, electronic media (CD-ROM), and print. The success of this experience gave Murray the confidence that the SEESAU project was achievable. UGA has a dedicated philosophy to providing patrons with complete access to the collection, whether print or electronic; • TEAkit: In order to improve communication with both public services departments and library patrons, UGA is in the early planning stages of creating a ticketing system for access problems called TEAkit. Essentially, a link to an electronic access problem reporting form will be included on all Web pages that link to an electronic resource. This form will allow patrons to submit access problems and receive feedback on the status of any access problems previously submitted including when the title is fixed. Potentially, once developed, this system may be merged with SEESAU's Problem Tracking Utility.
pdf. Murray, the developer and copyright holder for the SEESAU interface, is open to sharing his advice, knowledge and programming behind the interface, as well as providing local demonstrations. He can be contacted via email at
[email protected]. Other members of the SEESAU team who can be contacted for further information include Chandra Jackson (
[email protected]), responsible for reviewing and filtering reported problems; Laura Heilman (lheilman@uga. edu), SFX liaison and primary contact for problem resolution; and Dana Walker (
[email protected]), department head and project coordinator. For librarians interested in a system like SEESAU, what should they keep in mind? SEESAU itself is the Web interface or face of the results pulled from the ILS. Data extracted from any ILS used to drive SEESAU would have the same formatting requirements. In addition, a predictive system is required to schedule when the titles would appear on the report. Obviously, any data that arrives into the SEESAU interface as expected will work well with the system. Also, keep in mind that the terminology used within the SEESAU system is specific to the Voyager and SFX systems feeding the tool. If a library uses another link resolver besides SFX, links would need to be constructed to work and resolve within that particular system. Even if a library uses the same ILS and link resolver, SEESAU was designed to be used specifically with UGA's processes within Voyager. Some customization would be required to adjust SEESAU to local practices.
Sharing SEESAU
Many librarians have been reticent to pursue e-journal access verification for numerous reasons. When e-journals first became available, there was a false sense of security about the stability of the Internet. It did not take long, however, for perpetual access
The Future of Access Verification Systems
For readers interested in learning more about SEESAU, additional information is available at http://www.hindcite.com/SEESAU_notes.
86
Collins / Serials Review 35 (2009) 80–87
Figure 8. Archival Access Verification main page.
issues (or lack thereof) to arise. Even as librarians began to acknowledge that some level of access verification would be desirable, the growing volume of electronic journal titles to verify was overwhelming. When this task was weighed against other eresource tasks, such as licensing, registering access for new titles and knowledgebase management for link resolution and A-to-Z listing services, the return on investment seemed questionable. Thus, the fix-upon-failure philosophy has come to dominate the academic library landscape. SEESAU addresses several of the stumbling blocks that may have kept some librarians from pursuing a proactive approach to e-journal access verification. The process is systematic, requires minimal maintenance and, most importantly, it is quick and easy. For libraries that have an increasing number of electronic only journal titles, ensuring that access is available may become a greater priority.
Murray foresees that the future of access verification systems like SEESAU will be automation. Automated spiders or robots could do the access checking for a library leaving the library staff to focus on problem resolution. This level of automation would make it possible for searching to take place twenty-four hours a day or at least during the times determined most useful by library staff. Within five to ten years, Murray predicts that access verification systems will be included within the ILS or ERMS suite of modules or tools, very similar to the introduction of serials check-in modules within the ILS environment over the last few decades. Publishers may also take on this role and provide their own access verification systems as it is to their advantage to ensure that libraries have reliable, timely access to their journal titles. Ultimately, a systematic approach to access verification is to the library's advantage as well if it assists in better serving the needs of patrons.
87