Filip Radlinski has earned a two-year Microsoft Research Fellowship, awarded to outstanding Ph.D. students who are studying in either computer science, electrical engineering, or math departments. He will use the fellowship to further his thesis research, which involves applying machine learning to ranking problems. The fellowship covers a variety of expenses, including tuition, fees and conference travel and provides a stipend and tablet PC.
"The competition for Microsoft Research Fellowships is intense," said Computer Science Professor Thorsten Joachims who serves as Radlinski's advisor. "The fellowship is a tremendous asset to both Filip and to Cornell in how it recognizes the quality and the potential impact of his work on making search engines learn improved ranking functions."
Radlinski's research interests include using implicit feedback data to automatically improve search engines, adapting the benefits of large margin methods, such as support vector machines, to new tasks, using the tremendous amounts of data on the Web to better deal with text in traditional machine learning settings, and computer-assisted learning.
In 2005, Radlinski earned a Best Student Paper Award at the 11th annual international conference on Knowledge Discovery and Datamining for "Query Chains: Learning to Rank from Implicit Feedback," co-authored with Joachimes. In addition, he is a Fulbright scholar from Australia, president of the Australians and New Zealanders at Cornell, co-chair of the 2006 North East Student Colloquium on Artificial Intelligence (a student-run conference for AI students), and represents graduate students on Cornell's Computer Science Computing Facilities Support advisory committee.