---

Ratings

Methodology

A third frequently used method for personalized music recommendation engines to learn a listener's preferences is to have the listener manually input his/her ratings of the tracks, albums, and artists he/she knows.

Pros and Cons

A ratings system is the most accurate way for a personalized recommendation engine to learn a listener's taste in music, and this system has been successfully implemented by LAUNCHcast. However, the primary downside with this system is that it is time consuming for the listener, as many listeners who are not music fanatics are not willing to take the time to rate each track, album, and artist they know. Also, listeners are likely to give music they dislike after a few listens a negative rating, even though they might have just had a bad experience. Therefore, recommendations systems should incorporate some information about the listener's confidence in his/her ratings.