Netflix's $1 million contest to design a better movie preference-matching algorithm ended on July 26th, with two teams in a near tie and the outcome of the contest still in doubt.
Last month, an international team called BellKor's Pragmatic Chaos declared that it had developed algorithms that could improve the movie recommendations Netflix provides to its users by at least 10 percent. The contest rules then gave other teams 30 days to try to beat BellKor's. Minutes before the deadline, a team consisting of other reassembled teams, called The Ensemble, submitted an entry that just beat out BellKor's on the Netflix public Web leaderboard. However, BellKor's team members claim that they are the winners since Netflix contacted them to validate the entry, and that their entry had a better test score than The Ensemble.
Netflix's Steve Swasey says a winner has been named, and one likely will not be announced until sometime in September. He says the Web leaderboard is based on what the teams submitted, but Netflix's in-house researchers and outside experts must validate the teams' submissions and review the submitted code, design documents, and other materials.
From The New York Times
View Full Article
Abstracts Copyright © 2009 Information Inc., Bethesda, Maryland, USA