Building the digital library
Photo by Dan Addison
|Thornton Staples, director of U.Va.’s Digital Library Research and Development Group
By Charles Feigenoff
Things were much simpler for librarians when all that libraries had to worry about were books and journals.
Now their collections include videos, images, audio recordings and data sets. The great project that libraries have before them is not simply translating these disparate holdings into digital format and putting them online, though that is a significant task in itself.
More broadly, the challenge is to create a system for managing and organizing digital collections of materials in different media, and to do so in a way that allows users to take full advantage of the creative and scholarly options made possible by digital formats.
With support from the Andrew W. Mellon Foundation, the University Library is taking a leading role developing this framework. Members of U.Va.’s Digital Library Research and Development Group, working with the Digital Library Research Group at Cornell University, are creating the first digital object repository management system based on Fedora.
Fedora, which stands for flexible, extensible digital object repository architecture, is a trademarked series of standards for building repositories of all sorts, much like building codes set specifications for electrical, plumbing and heating systems that can be adapted for everything from a factory to a residence.
“We began by looking for a commercial digital library system, but couldn’t find one that met our needs,” said Thornton Staples, director of U.Va.’s Digital Library Research and Development Group. Cornell had published a paper on Fedora and developed proof-of-concept software that looked promising.
The U.Va. team reinterpreted the architecture and created a basic implementation that could support a very large-scale digital library. “We proved the concept, though it was clear that Fedora needed an industrial-strength management system,” Staples said. “At that point, we joined forces with them.”
U.Va. and Cornell are working together to create just such a system. At the end of three years of development, the open-source software has been downloaded more than 10,000 times by academic, government and commercial institutions in more than 50 countries. In the spring semester of 2005, the library began working with several faculty members to test their first implementation of the digital library system.
An important characteristic of Fedora is that it allows Staples and his colleagues to include detailed administrative and descriptive information with each digital object in their collection, plus links to software tools and services that have been configured to deliver the content, and the ability to collect references to related digital objects. This is a particularly necessary characteristic of a digital repository because there is no way of determining what a digital file represents by simply looking at its code.
For instance, a digital image of a Monet painting might be accompanied by information about access restrictions and its current location, a description of the image itself, tools to access the software needed to view it, and references to books in which it appears as well as to similar objects.
Staples is particularly enthusiastic about this last category of information because it gives users the ability not simply to search the collection, but to browse.
“In essence, we are giving users the ability to create networks of objects,” he says. “As new users come along, they can use the objects in different ways and create a new set of relationships that is then available to the next user. In essence, this is what scholarship is about.”
Reprinted from the spring 2005 issue of Explorations.
This issue highlights the stories of six U.Va. scholars who are publishing and exporting digital tools as part of their scholarship. For more, please see www.virginia.edu/researchandpublicservice/publications/explorations.html.