After more than a year, Netflix Inc. is still waiting to award $1 million to a design team that could help it build better personalization technology. But a team of AT&T Labs engineers has come close.
After more than a year, Netflix Inc. is still waiting to award $1 million to a design team that could help it build better personalization technology. But a team of AT&T; Labs engineers has come close.
In October 2006, Netflix Inc., No. 18 in the Internet Retailer Top 500 Guide, announced an award of $1 million to the first person or team that can help Netflix achieve certain accuracy goals in recommending movies based on customers’ personal preferences. To assist contestants in their efforts, Netflix has made available 100 million movie ratings logged by its subscribers; it has de-identified the ratings-which range from one to five stars-to keep subscribers’ data and opinions confidential.
After a year there’s still no big winner, but Netflix did award its first progress prize of $50,000 to "KorBell," a group of researchers at AT&T; Labs. The KorBell team, which works on visualizing and analyzing large networks for AT&T; in Florham Park, NJ, improved upon the Netflix recommendation system by 8.43%, Netflix says.
KorBell led a field of more than 27,000 contestants on more than 2,550 teams from 161 countries, says Netflix. "It was a closely fought race for the Progress Prize right to the wire," says Netflix vice president of recommendations systems Jim Bennett. "KorBell`s leading submission was posted to the Netflix Prize web site just 30 minutes before the deadline."
The threshold required to win the Netflix Prize is a 10% improvement in accuracy over the current Netflix recommendation system. Netflix will award a progress prize annually until someone wins the grand prize. “The grand prize of $1 million is still up for grabs,” Netflix says.