Blog Post

Netflix Prize’s Next Challenge: Demographic and Behavioral Recommendations

Stay on Top of Enterprise Technology Trends

Get updates impacting your industry from our GigaOm Research Community
Join the Community!

Three years after the competition started, the first Netflix (s NFLX) Prize was awarded today to “BellKor’s Pragmatic Chaos,” a seven-member squad that combined three competing teams (and included two AT&T (s T) scientists, the company is touting today).


And after all that, the winning team only won by 10 minutes, as reported after attending the award ceremony in New York City Monday morning, where seven winners met in person for the first time.

Amazingly — considering the length of the contest — the decision came down to a matter of minutes, according to Netflix Prize chief Neil Hunt. BellKor’s Pragmatic Chaos submitted their solution ten minutes earlier than the second-place team, The Ensemble, and the two teams’ algorithms were a perfect tie.

Netflix was able to wring three years of research to nudge its recommendation algorithm up 10.5 percent, at a cost of $1 million in prize money — a stunning feat on its own. Now the company is launching a new competition, and making its goal a little more attainable, by announcing it will award $500,000 to the top team six months into a new contest, and $500,000 to the top team at the end of 18 months.

The first Netflix Prize centered on a data set of users who have reviewed an average of 50 movies. Netflix Prize 2 “focuses on the much harder problem of predicting movie enjoyment by members who don’t rate movies often, or at all, by taking advantage of demographic and behavioral data carrying implicit signals about the individuals’ taste profiles.” Netflix is providing more than 100 million anonymized data points including age, gender, ZIP code, genre ratings and previously chosen movies. (Hey Netflix and contestants, let’s just get this out there in the open: I hate chick flicks. Don’t try to tell me I don’t.)

While some watchers had complained that the first Netflix Prize results were really not that extensible to situations outside of Netflix, the promise new research into demographic and behavioral correlations may help them change their tune. (And, to be fair, AT&T says it will use its researchers’ work from the first contest.) Though I have to say I’m a little surprised Netflix thinks it can get away with designating only another million dollars in prize money!

Photo courtesy AT&T.

5 Responses to “Netflix Prize’s Next Challenge: Demographic and Behavioral Recommendations”

  1. Here’s an interesting interview with the “Pragmatic” laymen sub-team that I published (if the URL link below is missing, find it on my Predictive Analytics World blog).

    Casual Rocket Scientists:
    An Interview with a Layman Leading the Netflix Prize, Martin Chabbert

    A couple of non-analytical laymen launched a mission to win the Netflix Prize, arguably the most high profile analytical competition to date. And these “casual part-timers” have succeeded – they compose one of three teams that together won the Prize. Martin Chabbert provides a tell-all interview, advocating adept engineering over innovative science. Their work and this competition as a whole are so cutting-edge, they make the space race look like good ol’ “Flash Gordon”.

    For the full article, see: