Out of thousands of submissions from teams from 186 countries, a winner has been declared in the $1 million Netflix contest, which asked competitors to improve its movie recommendation system using a data set of 100 million movie ratings by 10 percent or more. The challenge began in October 2006, and the winners are a seven-person team of statisticians and computer engineers who merged to pass the 10 percent threshold of the contest in June. They submitted their proposal 20 minutes before a competing team who also just barely made the 10 percent mark. The two were at a dead tie, but BellKor took the $1 million prize since they finished first. Netflix declared a new contest to model individuals' "taste profiles" from a data set of more than 100 million entries of ages, gender, ZIP Codes, and movie ratings. Half a million dollars will be awarded to the leading group after six months, and another half million to the leader after 18 months.