Yahoo! Selects Four CMU Grad Students as Future Thought Leaders in Computer Science

Yahoo! has named four Ph.D. students in Carnegie Mellon University's School of Computer Science among the 20 students selected as winners of its inaugural Key Scientific Challenges program, which recognizes outstanding graduate-student researchers with the potential to become thought leaders in their fields.

Pinar Donmez of the Language Technologies Institute (LTI) and Yi Zhang of the Machine Learning Department (MLD) were cited in the Machine Learning and Statistics category, while Jaime Arguello of LTI and Polo Chau of MLD won recognition in the Search Technologies category. No other university had as many winners in the program as Carnegie Mellon.

"We received an overwhelming number of outstanding applications and the competition in this first year of the Key Scientific Challenges program was very keen," said Ken Schmidt, Yahoo! director of academic relations. "We clearly were impressed by the quality of the applicants from Carnegie Mellon and believe Pinar, Yi, Jaime and Polo each hold great potential for making significant contributions as researchers."

Each recipient receives $5,000 in unrestricted seed funding for their research, exclusive access to certain Yahoo! data sets and the opportunity to collaborate directly with Yahoo! scientists. This summer, they will attend a Yahoo! Graduate Student Summit where they can present and discuss their work with some of the top minds in academia and industry.

Applicants to the program each outlined research efforts that addressed one of five key areas: community infrastructure and information management; computational advertising; economics and social systems; machine learning and statistics; and search technologies.

Donmez focuses her research on proactive learning, in which a program must evaluate information obtained from multiple sources that may not be equally reliable or available and may have varying costs. In particular, she is developing a framework that would choose the most cost-effective information elicitation strategy to help improve learning.

Zhang's research concerns how models assembled by statistical machine learning algorithms, such as a profile of a person's interests based on Internet activity, can be compressed into a concise, more usable form.

Arguello studies how to combine content from specialized search engines, such as for news, travel, or images, and in his most recent work, is investigating how to determine which of these search engines is most likely to contain information relevant to a given query.

Chau conducts research on integrating data mining and human-computer interaction to enable interactive mining of large graphs, such as social networks. His system helps analysts explore, visualize and understand large graphs, and pinpoint patterns, anomalies and interesting properties among them.

More information about the Yahoo! Key Scientific Challenges program is available at http://research.yahoo.com/ksc.