Talk:MovieLens/Old

Untitled
MovieLens is a research site run by GroupLens Research in the computer science department at the University of Minnesota. MovieLens uses collaborative filtering technology to recommend films to users. The predictions are personalized to individual user's tastes, requiring users to rate films they have seen and generating recommendations based on patterns of similarity discovered in the user base.

The original MovieLens algorithm ("user-user") generated correlation coefficients between all users. User recommendations were assigned a score based on their ratings by other users who had a high correlation coefficient. Recommendations were based on titles that were highly rated by users with similar movie preferences, or movies that were rated low by users who dislike the user's highly-rated movies.

MovieLens currently uses a tweaked version of a different published collaborative filtering algorithm ("item-item").

In 2006, it introduced folksonomy tags and forums.

Unlike other communities where ratings are public, Movielens keeps data about users secret as per academic experiment policy.