A $1-million prize offered by Netflix – an online movies rental service – has been claimed by a team of algorithm developers. Netflix had offered the prize three years ago to anyone who could improve the accuracy of its recommendation engine by 10%.
Netflix, like many other successful online businesses, relies on an algorithm to predict its customers preferences and recommend other movies for rental. Well over 50,000 contestants have attempted to rise to Netflix’s 10% challenge over the past three years. Two teams finally managed to improve Netflix’s recommendation engine by an identical 10.06%. Team Bellkor – consisting of seven researchers who had pooled their resources after working independently – managed to pip its rival team Ensemble to the post by a mere 10 minutes!
Update:
As a follow-up to this hugely successful prize challenge, Netflix has announced a second contest with another million dollars up for grabs. This time, they are encouraging developers to optimize software to predict a user’s movie preferences based on demographic information such as age, gender & domicile. Check out the Netflix Prize site for more details.
The New York Times also ran an interesting article on the emergence of the “prize economy” where organizations of various stripes are offering multimillion dollar prizes for advances in a range of fields.