Center aims to improve data mining

June 20, 2006

University of Illinois researchers will help lead an effort to improve techniques for pulling useful information from vast amounts of data, including information that may be important to national security.

Federal and UI officials this week announced that the UI will receive a three-year, $2.4 million grant for the project through the Department of Homeland Security's Institute for Discrete Sciences.

The grant will establish a new "Multimodal Information Access and Synthesis Center" at the UI to "advance the understanding and technologies required to deal with large amounts of information available today in multiple textual forms," said UI computer science Professor Dan Roth, who will direct the new center.

Besides being potentially useful to intelligence analysts, the data-mining techniques developed at the center also should be useful to scientists, businesses and individuals who need to make sense of incredible amounts of information available to them today, from customer buying trends to the human genome.

The grant was announced in press releases by federal officials, including U.S. Rep. Tim Johnson, R-Urbana, and the UI. Chancellor Richard Herman said the university's national leadership "in using technology to make information accessible and understandable" was a factor in obtaining the money.

Marc Snir, head of the UI Computer Science Department, noted that search and data-mining research already has led to a number of startup companies in the UI's south research park.

Other UI computer scientists involved with the project include Jiawei Han, David Forsyth, Kevin Chang, Cheng Zhai and AnHai Doan.