McCallum named to tenure-track faculty

After a national search for new faculty, the Department recruited a professor from within its own ranks. Andrew McCallum, who has been a Research Associate Professor in the Department since 2002, begins this fall as a tenure-track Associate Professor.

“I’ve had a wonderful first year advising students and doing research at UMass, but I also love to teach. I’m now looking forward to deeper involvement in teaching here as well,” says McCallum.

His particular interests include natural language processing, information extraction, Web mining, and finite-state models, all explored with the tools of statistical machine learning. Since arriving, McCallum has taught seminars in “Statistical Information Extraction” and “Computational Social Network Analysis.” In the future he will also be teaching courses in natural language processing at the undergraduate and graduate level.

During the past year, McCallum was affiliated with the Department’s Center for Intelligent Information Retrieval (CIIR) and the Knowledge Discovery Laboratory (KDL). He also collaborated with faculty in a number of other areas. Having expanded from advising only one student when he arrived at UMass, McCallum now leads a research team of seven graduate students, one undergraduate, and a senior researcher.

McCallum has been focusing on a number of new initiatives while at UMass. He is part of an SRI International research team developing a “Cognitive Agent the Learns and Organizes” (CALO), funded through DARPA’s IPTO. He is also working with Aptima, Inc. on a DARPA STTR-funded project to automate the diagnosis of usability reports. As PI with Research Assistant Professor David Jensen, McCallum was recently awarded an NSF Medium ITR to explore unified statistical models of information extraction and data mining. Their new research will be demonstrated in the construction of a public Web portal of scientific research and the social network of researchers. This past year McCallum and his students have also created MALLET, a Java toolkit for machine learning applied to natural language, which is now in use at CMU, MIT, UPenn, and several other universities.

Before joining UMass in 2002, McCallum was Vice President of Research and Development at WhizBang Labs and directed the company’s 30-person research lab in Pittsburgh, PA. He also held an adjunct faculty position at Carnegie Mellon University, where he co-advised several Ph.D. students and helped teach courses. Prior to WhizBang, he was a Research Scientist and Coordinator at Just Research (Justsystem Pittsburgh Research Center), where he spearheaded the development of methods for statistical text processing. McCallum was a post-doctoral fellow at Carnegie Mellon after receiving his Ph.D. in computer science from University of Rochester in 1995. His Ph.D. work, which was nominated for the ACM Best Thesis Award, did not focus on text, but rather on models of short-term memory and selective perception for reinforcement learning agents and humans.