Netflix Ditches Beat-Our-Recommendation-System Contest Due to Privacy Questions

Netflix's grassroots competition to best its recommendation algorithm was a bit popularity booster for the company--but thanks to concerns over privacy (backed up by one troubling example), the company has decided to kill the contest.

Netflix's recommendation engine can be a little wonky. (I know the film Big Fan is about football, and stars comedian Patton Oswalt, but it's also a painful drama written and directed by the guy who wrote The Wrestler, and it's an exceptionally bleak and dark story--so maybe Bring It On isn't the right movie to recommend, eh Netflix?) So they started a contest in which any development team that could create a movie recommendation algorithm more accurate than theirs would will a million dollars. It's a win-win-win: the company gets press, grassroots developers get a chance to win a million dollars, and we all get a better recommendation system.

Unfortunately, Netflix also had to release certain data to these teams for testing, as simply part of the process--and one woman sued for a violation of privacy. The woman was gay, and not broadcasting it, and felt that if her choice of films was made public, she'd be unable to retain her privacy.

The Federal Trade Commission investigated the situation, and Netflix settled the lawsuit for an undisclosed sum. Unfortunately, the settlement included a section banning Netflix from releasing any more of this data--which means the end of the competition. It's unfortunate, but probably a necessary evil. Netflix will just have to improve their algorithms on their own, now.

[Netflix via PCMag]

Add New Comment

1 Comments

  • Phoebe Spanier

    The Netflix competition was a great source of innovation, and from that perspective, it's too bad that the second competition can't continue. Improvements aside, it's true that Netflix recommendations sometimes fail to take into account mood and tone, as in the example cited.