KU Transforming Digital Library into Site to Aid Computer Scientists

University of Kansas researchers at the Information and Telecommunication Technology Center are collaborating on a National Science Foundation project to transform a digital library into a Web site to assist computer and information scientists. The Web site will sift through thousands of scholarly articles and highlight relevant text and citations, freeing scientists from much of the tedious legwork.

The NSF Computing Research Infrastructure grant supports the expansion of the Scientific Literary Digital Library (CiteSeer). The Web site receives more than 1 million hits a day from computer and information scientists and other researchers. Maintained by Penn State University, the CiteSeer searchable archive houses nearly 700,000 articles, primarily on computer and information science. The Next Generation CiteSeer project is adding new and expanding existing services, moving it from a static collection of information to a dynamic, collaborative research assistant.

"It is very exciting to do research on a high-visibility project that makes thousands of other researchers more effective," said Susan Gauch, director of ITTC's Intelligent Systems Laboratory and professor of electrical engineering and computer science in the KU School of Engineering.

Gauch is KU's lead researcher on the multi-institution project, lead by Penn State.

For CiteSeer, she will help create an interactive site that alerts researchers to newly published articles of interest and performs automated personalized searches. The increased reliability and sustainability will promote collaborative usage, leading to the formation of active CiteSeer research communities. The Next Generation CiteSeer will improve the ability of researchers, especially those who lack quality libraries, to quickly and effectively use the literature in their field.

Since developing one of the original Internet "metasearch" engines in 1993, Gauch has garnered international recognition for her online information retrieval research. Her search engines simultaneously examine multiple search engines, such as Yahoo and Google, for results. Gauch recognizes not everyone who types in a particular search phrase is looking for the same results. She recently developed filters to help people tailor their online searches for medical information. Users can select the qualities that matter to them, such as depth of information, privacy and timeliness.

Search engines usually return more than 1,500 results per inquiry. Of the top 20 results, often only half will be relevant to the user, according to Gauch. Her technology greatly improves those odds.

"CiteSeer is a tool used by many computer scientists. Therefore, the results of Dr. Gauch's work will have a broad impact for the research community," said Victor Frost, ITTC director and Dan F. Servey distinguished professor of electrical engineering and computer science.