Netflix may have improved movie-ratings system
PHYSICAL: BellKor claims to have 10% solution for recommendation engine
By Danny King -- Video Business, 6/29/2009
JUNE 29 | PHYSICAL: More than two years and 49,000 submissions after Netflix introduced a contest geared to improve the accuracy of its user-generated movie-recommendation system, one group is claiming the grand prize.
A team that includes AT&T and Yahoo! Research says it should win the $1 million grand prize that’s supposed to be awarded to the first group of programmers that can improve Netflix’s movie-recommendation engine by more than 10%. The group, named BellKor’s Pragmatic Chaos, put together a program that improves Netflix’s rating service by 10.05%, Web site HackingNetflix.com said late last week.
The improvements are part of a contest sponsored by the largest U.S. movie-rental service via mail that was started in 2006 and expires in October 2011. Netflix paid out $50,000 in progress prizes for both 2007 and 2008, but no team has claimed the grand prize.
Netflix’s movie-recommendation engine processes more than 2 billion user ratings in order to make recommendations for a particular subscriber based on how that subscriber has rated other movies.
Despite the BellKor team’s claim, Netflix will need about a month to verify whether the BellKor team wins the prize, company spokesman Steve Swasey said. During that time, other teams can continue to submit claims on the prize in case BellKor’s solution can’t be verified, according to Swasey.
