http://www.nytimes.com/2008/11/23/magazine/23Netflix-t.html?pagewanted=all
This week’s New York Times Sunday magazine checks in with a few competitors in the Netflix competition. One of the contestants estimates that an accurate predition of whether someone would like the movie Napoleon Dynamite—a film that viewers seem to either love or hate—would put him 15% closer to the algorithm that would earn a $1 million prize.
When Bertoni runs his algorithms on regular hits like “Lethal Weapon” or “Miss Congeniality” and tries to predict how any given Netflix user will rate them, he’s usually within eight-tenths of a star. But with films like “Napoleon Dynamite,” he’s off by an average of 1.2 stars.
According to the article, the “Napoleon Dynamite problem” exposes the “a serious weakness of computers”: their inability to anticipate all of the factors in a person’s decision-making process. Someone could decide to watch a movie after a Blockbuster clerk’s passionate recommendation, or to understand a cultural reference point, or simply to try something different.
Another critic of computer recommendations is, oddly enough, Pattie Maes, the M.I.T. professor. She notes that there’s something slightly antisocial—“narrow-minded”—about hyperpersonalized recommendation systems. Sure, it’s good to have a computer find more of what you already like. But culture isn’t experienced in solitude. We also consume shows and movies and music as a way of participating in society. That social need can override the question of whether or not we’ll like the movie.
An interesting read. [If You Liked This, You’re Sure to Love That | NYTimes]
Comments [0]
Comments [0]