The project is directed by Manuel Pérez-Quiñones, Virginia Tech assistant professor of computer science. Other team members are: Weiguo Fan, assistant professor of accounting and information systems, and Edward Fox, professor of computer science, both at Virginia Tech, and Lillian Cassel, professor of computing sciences at Villanova University.
"Our goal," Pérez said, "is to get content from the National Science Digital Library (NSDL) closer to its intended audience." The project's target beneficiaries are students and professors in all areas of computing. "We hope to extend the library's reach into the educational system and thus increase its number of users."
The NSDL has a vast collection of resources for education and research in science, technology, engineering, and mathematics. Its Web site notes that in contrast to a single-term Google search that will produce millions of results, most of which may be irrelevant to science and technology education, NSDL looks for only resources drawn from "high-quality providers of scientific and technical materials that are accurate and appropriate for educational settings."
As its collections grow, the NSDL is placing more emphasis on user services and higher-level functionality, Pérez said. But in helping to expand the library's usefulness, he adds, researchers must not assume that a "feature-rich" Web site -- with browsing, searching, recommendations, and discussion forums -- would be enough to draw users. That would be like "building a large library with lots of facilities, but placing it at the outskirts of campus -- few teachers and students will interrupt their daily routine to use the new facilities."
The project focuses on course Web sites, he said, as they are "the most commonly used application in today's educational environment, typically providing access to a course syllabus, schedule, assignments, grades, and discussion forums." The technology that the researchers will develop will allow a course Web site to be the entry point into the NSDL's collections. The product would be "a flexible and personalized information interface that supports both exploratory and focused searches and provides the ability to obtain context-sensitive services at various levels of interaction."
Explaning the need for "personalizing" the library interface, Pérez said the majority of NSDL collections do not gather information about their visitors that would allow them to match resources to the user's requirements. As for "context-sensitive services," he cites an example in which one professor visiting an NSDL collection might be looking for a list of compiler textbooks in order to determine the most common programming languages used in compiler instruction. Another professor might seek the same textbook list, but for the purpose of assigning readings. For this professor, a link would appear on an NSDL-enabled course Web site that allows him or her to see "other textbooks appropriate for this course."
Such "context-sensitive" service, he said, can result in a better use of NSDL resources. "Users are more likely to select options right at the spot and context where they are doing their work, instead of going to a different Web site, searching for textbooks, and browsing through the list of textbooks to identify those that might be appropriate for their needs."
The project will include user studies. "We have conducted several workshops and have learned that the social dynamics of online sharing of educational materials are very complex -- going beyond 'build a website, and people will use it.'" He said that in one workshop participants reported that they routinely looked for online resources to enhance their classes. Only about 40 percent said that they would share or communicate any modifications or adaptations they made to a resource with the original author, while about 70 percent said that they would do so if there were an online community for this purpose.
For the NSDL and its collections to have a broader reach, he said, more knowledge is needed about how users interact with digital libraries.
The project is part of continuing research efforts at Virginia Tech and Villanova on digital libraries, personalization, and human-computer interaction.