Share and Share Alike
The eScholarship Repository, a project of CDL's eScholarship initiative that launched in April 2002, offers faculty a central location for depositing any research or scholarly output deemed appropriate by their participating UC research unit, center, or academic department. Scholars may upload their work at no cost; users may download that work free of charge.
Available at http://repositories.cdlib.org/escholarship/ the repository enables the rapid creation, management, and dissemination of journals, peer-reviewed series, working papers, technical reports, research results, and other forms of scholarship generated by UC researchers. "Anyone can submit a paper and anyone can access the information on the site at no cost, but only university faculty can have a site within the repository and determine what goes into it," she explains.
To date, nearly 130 units within the UC system have signed up to participate in the project, which is open to all disciplines ranging from the social sciences and humanities to business and education. "The model requires that a member of each organized research unit or faculty department serve as a gatekeeper of the information," Candee continues. "Each assigned administrator is trained in under an hour to upload material." The contents of the repository expand almost daily. In mid-November 2003, for example, the repository provided visitors free access to 2,366 papers; by mid-January, that number had grown to 2,451. Other searchable resources to be found within the repository include one journal, five edited volumes, and dozens of research series.
At the heart of the repository is a suite of digital publishing tools developed by Berkeley Electronic Press (bepress), which was founded in 1999 "to address the inefficiencies that characterize the current scholarly publishing model," says president Greg Tananbaum. "At our core, we have developed EdiKit, an end-to-end publication system that handles everything from submission to dissemination electronically." Among its services, bepress licenses EdiKit to third parties, providing back-office and electronic publishing services, and teams with universities to operate their institutional repositories. (Clients in this space include UC, Florida State University, and Boston College.)
"The thing that is most powerful about bepress is its ability to handle publishing" from beginning to end, Candee adds. "We experimented with different software packages for more than a year before we discovered bepress in 2001. Together, we modified and extended the capabilities of the EdiKit software to do what we needed to achieve."
Specifically, CDL wanted the software that drove the repository to comply with the Open Archive Initiative metadata harvesting protocol. "Any one of the papers within the repository can be found with Google," she says, giving users worldwide accessibility to the repository's contents via a number of locations and search methodologies. The software also allows each UC unit to have its own organizational identity within the repository, tracks usage of resources within the system, and makes all content fully searchable and persistently available.
Today, UC pays bepress to run the process on its server. "We outsource all technical support and production to them. We manage the people; they manage the process."
Nearly two years after its debut, the eScholarship Repository is already a success story in terms of the speed with which it has been embraced. "Today, we log more than 11,000 full-text downloads of papers per week," Candee says, noting that in its first year, the average was 2,000. "The barrier now is extending the technology to allow batch-loading and faster downloads."
To address this issue, CDL and bepress unveiled a feature in early 2004 that streamlines this process. "If the material in question is already in electronic form—as a Word document, PowerPoint presentation, or PDF, for example—the UC site administrator can simply point to the file's location on the server and Batch Upload will automatically write all of the metadata for him," Candee explains.
"Getting people to recognize the value of the system we've developed has been a huge challenge," she continues. To increase the repository's visibility, Candee and her staff regularly visit UC campuses to promote the program. Candee also has appointed liaisons at each campus who help train faculty and staff to work in the repository. bepress, meanwhile, recruits UC research units to participate, trains these groups to use it, and serves as a primary source of technical support.
Looking ahead, Candee says she would like to see the repository reach a wider audience. "Right now, our audience is any faculty member or researcher within the University of California. We may open it to other universities at some point." She also sees the repository expanding to include seminars and conference series. "That's already been approved, so we'll be adapting the software yet again to enable that."
Candee is adamant in her belief that the current publishing paradigm could use some tweaking. "Ideally, I'd like to see our model [of online publishing] become the standard in the academic world," she says. "I would like to see an alliance between digital libraries, university presses, and scholarly societies—the keepers of the scholarship record and the university's knowledge. In order to work together, we have to use standardized, interoperable tools and build the infrastructure to support it.
"The thorn is that so much scholarship output goes to commercial publishers and then we have to buy it back at huge costs," she continues. "We're being bled to death by them. So much money is spent on research, and then that stuff is given away for free to the commercial publishers and the libraries have to buy it back" to provide it to their users.
"We have to keep that knowledge in the university," she adds. "To do that, we have to develop the mechanisms to harness knowledge, innovation, and technology and get our educational and cultural resources under our control."
While the eScholarship Repository model may not be the be-all, end-all solution for every knowledge manager looking for a way to publish, archive, and search content, it's certain-ly a start.