Professor gets computing's `Nobel' ; Harvard's Leslie Valiant wins A.M. Turing Award

Harvard University professor Leslie G. Valiant, a pioneer in thefield of artificial intelligence, has been awarded the 2010 A.M.Turing Award, the most prestigious prize in the field of computerscience. Valiant's research into processes to make computers reasonas humans do laid the groundwork for applications ranging from e-mail spam filters to IBM's Watson computer system, which last monthbested human competitors on the game show "Jeopardy!"

Known in computing circles as the "Nobel Prize in Computing," theaward is bestowed by a scientific society, the Association forComputing Machinery, and includes a $250,000 cash prize funded byIntel Corp. and Google Inc.

The prize is named for Alan Turing, a British mathematician andcode breaker during World War II who is known as a father of moderncomputer science.

"This connection with the achievements of the previous winners,and of Turing himself, is more than anyone in my field canreasonably expect," Valiant wrote in an e-mail.

He will receive the award on June 4, at the association's annualawards banquet in San Jose, Calif.

"I think Les Valiant's award is one of the strongest TuringAwards ever," said Jennifer Chayes, a member of the selectioncommittee and the managing director of Microsoft Research NewEngland. "His work is absolutely amazing."

As an example, Chayes cited Valiant's 1984 paper "A Theory ofthe Learnable" as a landmark in the field, saying it led to theprocesses that allow computers to discern which e-mail messages areviewed as annoying, and which Web search results are most relevant.

"All of the personalization that we are seeing is due to thecomputer learning from what you did, and trying to find patterns inwhat you did," Chayes said. "That is absolutely an application ofhis research."

Valiant's work has "an adventurous aspect," said Jon Kleinberg, aCornell University computer science professor who endorsed theHarvard professor's nomination for the award.

"He takes on questions that are fundamental, but very hard toattack, like how do intelligent agents learn? Or how does the braincompute?" Kleinberg said.

The science of machine learning, the type of artificialintelligence work Valiant helped to pioneer, develops the tools bywhich computers can solve problems on their own rather than rely onpreprogrammed answers, Kleinberg explained. One real-worldapplication is speech-recognition software, he added.

"Machine learning has allowed computers to display that samelevel of subtlety and complexity, the way that humans do when theysolve problems," Kleinberg said. "You pick up the phone and call anairline, it's almost impossible to get a human being. You're talkingto computers."

Valiant's win brings the prize to Cambridge for the second timein three years. A Massachusetts Institute of Technology professor,Barbara Liskov, won in 2008. She was recognized for the developmentof object-oriented programming, a method used to write most of thesoftware programs that power personal computers and the Internet.

Calvin Hennick can be reached at calvinhennick@yahoo.com.

09turing.ART