Abstract
The speakers discussed practical steps that smaller institutions could take to develop their own institutional repositories (IRs) from scratch using local resources. Yuji Tosaka spoke about the overall context and problems that IRs can present for smaller institutions; Cathy Weng then gave an overview of The College of New Jersey's efforts to develop an IR and discussed the college's MUSE project, a college-wide mini-grant program for working with selected undergraduate students on research and creative activity, and how this internal opportunity led to an IR pilot project. Tosaka ended by speaking about life after MUSE and the still-ongoing IR development.
THE COLLEGE OF NEW JERSEY
The College of New Jersey (TCNJ) is a state college located in Ewing, New Jersey, about ten minutes from Trenton, the state capital, and Princeton. It is a highly selective residential college made up of seven schools that focuses on the undergraduate experience, with a full-time equivalency (FTE) of about 6,000 students. Faculty and undergraduate research, especially in collaboration, is strongly encouraged and supported. The library is medium-sized, with over half a million volumes. Currently, the library has few digital library collections.
INSTITUTIONAL REPOSITORIES
Tosaka began by defining an institutional repository (IR) as a digital library collection and service that is designed to manage, organize, and showcase the intellectual output of an academic community to a broader audience. IRs were started in the early 2000s, and since then have reached significant mass, with millions of downloads at many large repositories.
Large research universities were early adopters of IRs that focused on disseminating and preserving their faculty scholarship. Large universities still vastly outnumber smaller universities and colleges in IR development. However, as noted by the 2007 Census of Institutional Repositories in the United States, “there is still a “sleeping beast of demand for institutional repositories … from master's and baccalaureate institutions.” This report also found that librarians at such smaller institutions “want to learn about their peers’ experiences with IR costs, required technical expertise, funding the IR effort, whether the local learning community will contribute to and use the IR, and raising the issue of IRs with their institution's central administration.”Footnote 1
Tosaka next discussed the benefits of IRs, including the capability to allow universities and colleges to take stewardship of the intellectual output of the campus community by centralizing and providing long-term access to the institution's intellectual assets in ways that are not well supported by traditional library and publication models. IRs promote open access and dissemination, and increase the visibility and impact of faculty scholarship in the digital age. In addition, IRs can showcase student research and learning outcomes and thereby demonstrate the academic and educational quality of the institution. IRs are also important for institutional advancement and accountability as an official medium for communicating institutional accomplishments. In particular for public colleges and universities, they help enhance public accountability by showing a return on investment, that is, the results of publicly supported faculty research and education conducted on campus.
IR CHALLENGES AT SMALLER INSTITUTIONS
Perhaps the biggest barrier to IR development at smaller institutions is the problem of limited resources: funding, staffing, and technical expertise and support. At the TCNJ Library, Tosaka noted that there is only one systems librarian who takes care of the library information technology, whereas at a large research institution, there may be many library IT staff. Given these limitations, Tosaka and Weng felt that small institutions such as TCNJ need to pursue a minimal-cost approach to developing and maintaining IRs as new library initiatives.
IR IMPLEMENTATION OPTIONS AT SMALLER INSTITUTIONS
According to two recent studies that Tosaka cited, the predominant option for small institutions to implement an IR is through a consortium.Footnote 2 More than 60% took this route.Footnote 3 However, consortia options may not be available for all libraries. Outsourcing to vendor-hosted platforms like Berkeley Electronic Press (bepress) Digital Commons is another important IR choice, adopted by about 20% of small institutions.Footnote 4 However, although proprietary vendor platforms may be easy and quick, they are fairly expensive with annual license fees and may not be affordable. That leaves independently operated repositories as a third and least adopted option. This has been the IR route adopted so far at TCNJ.
IR IMPLEMENTATION AT TCNJ
The TCNJ IR initiative began in spring of 2009. A digital projects working group was formed at the library to investigate and identify library materials that should be digitized for preservation and broader public access purposes. Developing an IR was identified as one of the targeted services. Without a specific product, however, it was difficult to sell the idea of developing an IR to the library administration and to ask for financial and physical support. The group was also reluctant to pursue a commercial product as there were several open-source IR products that seemed promising. Despite these concerns, the Library Dean approved the initiative, and the group explored several open-source IR platforms, including DSpace, Greenstone, and Fedora. The group also seriously considered working with Rutgers University Library to host the IR using their home-grown digital repository system, RUCore, while developing TCNJ's own IR portal. However, none of the explored options proved feasible. As a result, the working group decided on a new strategy: that of a pilot IR, utilizing campus resources and with student involvement in the development, with the ultimate goal of having a permanent and sustainable library service.
TCNJ MENTORED UNDERGRADUATE SUMMER EXPERIENCE (MUSE) PROGRAM
The TCNJ MUSE program is a summer research program that lasts eight weeks from June through July, where undergraduate students conduct research or engage in creative activity in collaboration with TCNJ faculty members, who serve as mentors. Both students and faculty members apply for a research grant that is reviewed and selected by the Faculty-Student Collaboration Program Council through a highly competitive process.
LIBRARY MUSE IR PROJECT
The IR pilot project seemed to be a good fit for the TCNJ MUSE program. With this in mind, the library digital projects working group formed a library MUSE project team composed of three library faculty members—the Emerging Technologies librarian, the Head of Cataloging, and the Cataloging/Metadata Librarian—to handle various tasks, including technical needs, content recruitment, metadata application, rights management, and IR public interface customization. For the mentoring aspect of the project, the project team recruited two students majoring in computer science to learn about library repository systems, system installation and server administration, about working in an open-source community, and to help customize an open-source IR to meet local needs. The pilot project team submitted their MUSE application package in February 2011 and the team was awarded a MUSE grant of $9,975 in March 2011. This was extremely significant for the library and for the three library faculty involved, because it was the first ever library MUSE project and represented the library's active participation in academic mentoring. In addition, it recognized the library faculty as part of the research community. Most importantly, the success of the grant application indicated that the MUSE review council acknowledged the importance of having a campus repository system to store, preserve, and showcase the academic community's scholarly output.
MUSE IR IMPLEMENTATION USING IR+
After receiving the grant, the MUSE team critically and carefully evaluated open source IR platforms. Although at one point the team leaned toward DSpace, they ultimately decided to adopt IR+, a newer open-source system developed by the University of Rochester.Footnote 5 IR+ had several promising features, including the ability to browse by author, publication, and sponsor; faceted filtering; an author's workspace for collaboration and self-archiving; name authority control; and a profile page for each researcher.
TECHNICAL REQUIREMENTS
After deciding on IR+, the team established contact and closely communicated with the IR+ developer and repository coordinator, Nathan Sarr, while assessing the system's hardware requirements. They chose to have a physical server over a virtual server, in part because the IR+ manual was written for Windows server, but also in order to allow the students to learn server administration. By reviewing the installation, system administration, and user manuals, the team learned to use IR+ as administrators, users, and authors.
CONTENT BUILDING
To build content, the team reached out to both the library and the Department of Chemistry. At TCNJ, librarians have faculty status, so it was easy to obtain a list of publications authored by TCNJ Library faculty members. With the help of the Science and Engineering Librarian, the team also successfully recruited content from four Chemistry department faculty members. With these lists of publications, the team began to establish a preliminary metadata application profile (a local guideline that documents how information should be entered in each metadata element) and to create metadata.
COPYRIGHT MANAGEMENT
The team sought copyright clearance advice from colleagues in other institutions who had already implemented their IRs. Based on their colleagues’ advice, the team consulted Securing a Hybrid Environment for Research Preservation and Access/Rights Metadata for Open Archiving (SHERPA/RoMEO) to check publisher copyright policies on self-archiving for instructions on posting articles in an IR.Footnote 6 They then contacted publishers as needed to clarify copyright further, using e-mail to document the process, and established local profiles for individual publishers and journals in the IR.
LOGISTICS OF WORKING WITH STUDENTS
The team set up a Google group site to communicate with the student team members, and held weekly in-person meetings to discuss progress and assignments for the following week. The Emerging Technologies librarian met with the students almost daily, and because the students’ workstations were located in the Cataloging Department, Tosaka and Weng could hold informal discussions with the students as needed. Later on, the students joined the IR+ community and received technical support and guidance for coding from Nathan Sarr, the original software developer. In hindsight, Tosaka and Weng stated that it would have been more beneficial for the students to have received help with coding earlier on in the project.
PROJECT OUTCOMES
At the end of the MUSE program, the team successfully implemented TCNJ's pilot IR, dubbed TCNJ Digital Scholar. The team made local enhancements, including a more intuitive metadata creation process, and created over seventy records for articles, book chapters, PowerPoint slides, and poster presentations. In addition, they established a preliminary metadata application profile and also a preliminary rights management workflow. In addition, one of the student team members contributed coding to enhance IR+’s importing and exporting function for MARC 21 file batches. The new coding was then incorporated into IR+ 2.1, which was later released for use by the community.
LIFE AFTER MUSE
Following the end of the MUSE program last summer, Tosaka and Weng took several slow but steady steps to move the IR pilot project forward into becoming a sustainable library service, including securing support from the library administration and faculty, and also formalizing various policies and procedures. The Library Dean had been supportive of their efforts throughout the MUSE program. Following the end of the project, Tosaka and Weng approached him to discuss ways to move a new IR initiative forward in the library and obtain additional funding from the Academic Affairs department to hire a student worker to help with further IR development.
It was also vital to have faculty buy-in in order to proceed. In some ways, support from the library faculty was equally and perhaps even more important in TCNJ's organizational culture, where librarians have faculty status. Without faculty support, a new library-wide initiative could not succeed, even though the Dean had given his approval and resources for the project. To keep the faculty informed and to address any concerns, Tosaka and Weng provided the library faculty with regular status reports on the IR's progress, an IR demonstration, and a Q&A document. In the end, the library faculty expressed support for IR planning as a new library initiative. There is now a work-in-progress to develop a formal IR proposal, with input and comment from the entire library faculty, which will allow the Dean to take further action on the IR initiative.
In addition to gaining support from the library administration and faculty, Tosaka and Weng also continued to develop policies and procedures, including a metadata application profile to further prepare for a sustainable IR. Having a metadata application profile as a quality control mechanism was crucial to their goal of maintaining a sustainable IR with minimal costs. Since it would not be feasible for a professional librarian to work full-time on the IR, paraprofessionals and student workers would be tasked with creating and validating metadata records. Thus, it was necessary to provide clear guidance and documentation for them to follow. (More recently, Tosaka and Weng found that a student worker can handle routine metadata creation when good documentation is provided.)
In addition to metadata creation, copyright management was another area where clear policies and procedures were needed in order to maintain a sustainable IR. According to a recent study, libraries often struggle with keeping track of copyright permissions when using such tools as e-mails, hard-copy printouts, and spreadsheets.Footnote 7 In a bepress webinar in April 2012, one library reported that they had a librarian and a senior paraprofessional, assisted by student workers, who each spent 25% of their time on copyright management, using complicated in-house procedures.
Tosaka noted that since the TCNJ Library does not have such staffing available, they needed a simple workflow that could be managed by paraprofessionals and student workers. To that end, they experimented with using Centralized Online Resources Acquisitions and Licensing (CORAL), an open-source ERM system, to create a copyright workflow for a student worker.Footnote 8 Since their Systems Librarian and Electronic Resources/Serials Librarian had been working in the prior year on implementing CORAL, Tosaka and Weng had learned that CORAL could be a good tool for managing IR copyright workflow. In addition, using an in-house product was a big plus, given the constraints on funds.
CORAL is made up of four modules—Organizations, Licensing, Resources, and Usage Statistics—but it is not necessary to implement them all. Tosaka and Weng used only the Licensing module for the copyright workflow, and modified the status to display what publication version could be deposited in the IR (e.g., pre-publication or post-print) and they also adapted the expressions to hold related information and the copyright Uniform Resource Locator (URL), which they obtained from the SHERPA/RoMEO website.
WHAT IS NEXT?
One of the most critical next steps for the IR is content recruitment. Although Tosaka and Weng have not yet done much in this area since the MUSE pilot project, they expect that the TCNJ IR, like many IRs at smaller institutions, will likely focus on student work, reflecting the increasing emphasis at TCNJ on deep student learning and intensive faculty–student collaboration in scholarly and creative activity.
Another important step is campus outreach and buy-in. Tosaka and Weng are planning multiple approaches to promote the IR as a unique library service, including faculty mailings and presentations, individual subject librarians’ outreach, and partnership with other campus units. In addition, they plan to further refine their copyright management workflow. If CORAL works, they may suggest to the University of Rochester to incorporate it into the IR+ software for interoperability.
CONCLUSIONS
As Tosaka noted, open-source software is not free, as it requires server and staff time. However, even with limited local resources, the TCNJ IR initiative has shown that it is possible for a small group of committed librarians to create a foundation for establishing a sustainable, open-source IR at a small academic institution. In doing so, however, they often found that they had to think like a small startup or entrepreneur, by being flexible and trying any approach that might work, and by using whatever in-house resources they had, such as CORAL.
To illustrate what they felt was the ideal IR process map, Tosaka and Weng displayed an image from their college's strategic planning material, consisting of the phrase, “Formulate → Implement → Review and adjust.” Given the limited resources and staff at TCNJ, Tosaka and Weng did not aim for one big rollout. Instead, they quickly formulated a “good enough” plan and moved forward immediately to implementation. Then, they constantly reviewed their progress, adjusted their plan, and moved on to the next incremental cycle.
Last but not least, Tosaka and Weng cautioned against fearing failure. They noted that in smaller libraries, it is easy to feel that if something fails, there is little time, resources, or staff to take on a new initiative. But they noted that this is an entirely incorrect approach. They advised librarians at smaller institutions to keep in mind the larger mission and goal that they want to achieve. For example, if someone notices that the campus has IR needs and that the library has organizational and information architecture services that are unmatched on campus, then the correct questions to ask should be, “What's the road ahead? What resources do we have in-house? How can we get to our goal one step at a time?” In conclusion, Tosaka and Weng feel that by keeping their eyes on their ultimate goals and by exercising some creativity, smaller libraries can take on new initiatives with limited resources.
QUESTIONS AND ANSWERS
The first question asked during the question and answer session concerned TCNJ's use of CORAL and whether it was necessary to add journals to the system. Tosaka noted that CORAL comes pre-populated with many journal titles, and that it was easy to customize the fields to add information for the IR, including links to publishers. Another question regarded the number of libraries that develop IRs in-house. From their research, Tosaka and Weng found that most IRs are developed internally.
Several questions were raised about the technical nature of maintaining the IR. In response to an audience member who asked if there were staff members available with technical expertise, Tosaka noted that although the Emerging Technologies and Systems librarians have such expertise, they do not have any coding experience, which was what was needed to develop and maintain the IR. This raised a question of whether anyone understood how to operate the backend server, to which Tosaka responded that the student workers documented everything before they left, and that the Emerging Technologies Librarian would handle any upgrades in the future.
A cataloger asked if they had integrated the metadata in IR+ into the Online Public Access Catalog (OPAC). Although they have not yet, both felt that this would be a good idea, and Tosaka noted that batch loading from IR+ is now available. Finally, an audience member asked how well IR+ interacts with the library's discovery layer, to which Tosaka replied that the discovery layer can retrieve data from the IR.
Notes
1. “Census of Institutional Repositories in the United States,” Council on Library and Information Resources, last modified 2007, http://www.clir.org/pubs/reports/pub140/pub140.pdf, 74–75 (accessed November 19, 2012).
2. Jingfeng Xia and David B. Opperman, “Current Trends in Institutional Repositories for Institutions Offering Master's and Baccalaureate Degrees,” Serials Review 36 (2010): 10–18.
3. Melissa Nykanen, “Institutional Repositories at Smaller Institutions in the United States: Some Current Trends,” Journal of Electronic Resources Librarianship 23 (2011): 1–19.
4. Ibid.
5. “IR+ Institutional Repository Software,” University of Rochester, 2012, http://code.google.com/p/irplus/ (accessed November 19, 2012).
6. “SHERPA: Securing a Hybrid Environment for Research Preservation and Access,” http://www.sherpa.ac.uk/ (accessed November 19, 2012); and “RoMEO: Rights Metadata for Open Archiving,” http://www.sherpa.ac.uk/romeo/ (accessed November 19, 2012).
7. Ann Hanlon and Marisa Ramirez, “Asking for Permission: A Survey of Copyrights Workflows for Institutional Repositories,” portal: Libraries and the Academy 11 (2011): 683–702.
8. “CORAL,” University of Notre Dame Hesburgh Libraries, http://erm.library.nd.edu/ (accessed November 19, 2012).