Doug Downey, assistant professor of electrical engineering and computer science at Northwestern University’s McCormick School of Engineering and Applied Science, has received a Faculty Early Career Development Award from the National Science Foundation.
The program offers the NSF’s most prestigious awards for new faculty members. It recognizes and supports early career development of teacher-scholars who are most likely to become academic leaders. The minimum award size is $400,000 for a five-year period.
“The NSF CAREER award is an honor and great enabler for my group’s research,” Downey said. “The support will make it possible for us to deliver impactful research, train multiple Ph.D. students and provide new educational opportunities.”
Downey’s research interests fall into the areas of natural language processing, machine learning and artificial intelligence. He is interested in methods that learn automatically from large data sets and, in particular, new ways to extract knowledge from the Web to power new Web search capabilities.
The CAREER award will support research on how to integrate and scale Web information extraction (WIE) systems. Many different teams in academia and industry are pursuing WIE (Google’s Knowledge Graph and the open-source WikiData project, for example). Currently, different teams lack methods for combining their knowledge into a more powerful whole. Downey’s project will explore how to integrate knowledge automatically across WIE systems and approaches.
“Our goal is to make it possible for different knowledge bases — even if they are constructed very differently and use different internal representations — to exchange information automatically in natural language,” Downey said.
The CAREER award’s agenda includes the development of a related educational module to teach high school students how computers learn from data. The module will be made available to area high school students through Northwestern’s Office of STEM Education Partnerships, with a focus on reaching student groups currently underrepresented in computer science.