Two years ago, DVD rental service Netflix challenged the world to improve its movie recommendation system by 10%. So far, no one’s been able to do it: For the second year in a row, the company is awarding a $50,000 “progress prize,” but no $1 million grand prize.
Netflix: The team “BellKor in BigChaos” won the 2008 Progress Prize with a 9.44% improvement in the Netflix recommendation system. The team is actually a combination of two teams that previously competed against each other in their efforts to outpace more than 35,000 other teams from 180 countries vying for the $1 million grand prize.
A board of judges, including senior engineers at Netflix and well-known members of the machine learning community, validated BellKor in BigChaos’s entry as the winner of the second annual Progress Prize.
Last year the BellKor team, then known as KorBell, won the $50,000 progress prize with an improvement of 8.43 per cent. At that time BigChaos was in 81st place with a 5.51% improvement. BigChaos has since advanced to a leading position, vying with BellKor for 2nd place behind their winning combined solution.
The BellKor side of the team is comprised of three scientists, Chris Volinsky, Yehuda Koren, and Bob Bell, who until recently worked together at AT&T Labs in Florham Park, N.J.. Recently, Koren left AT&T Labs to join Yahoo Research in Israel.
BigChaos members Andreas Töscher, Michael Jahrer, Georg Preßler and Michael Schrotter founded Commendo Research in Köflach, Austria earlier this year based on the notoriety they received in the Netflix Prize competition.