---

Playlists

Methodology

A method often used by personalized music recommendation engines to learn a listener's preferences in music is to analyze the music in the listener's playlist(s).

Pros and Cons

Analyzing the music in a listener's playlist is a quick, easy, and convenient way to analyze a listener's preference in music, however it is far from accurate or scientific. Listeners will listen to some tracks, albums, or artists much more often than others, while never listenening to some, and this information should be taken into consideration. Furthermore, listeners might not have all of the music they know in their playlist, so part of their range of musical understanding would be overlooked.

Playcounts

Methodology

Another common method used by personalized music recommendation engines to learn a listener's preferences is to keep track of the frequency with which certain tracks, albums, and/or artists are played.

Pros and Cons

Although more complicated than playlist analysis, keeping track of playcounts is a considerably more accurate method of determining a person's true taste in music. However, analyzing playcounts has a few downsides. Listeners will often listen to a new track, album, or artist a few times before determing if they like it or not, and these plays will be automatically interpreted as a sign of affinity for the music. Also, if a user accidentally leaves his/her computer on playing music, this irrelevant data will also be taken into account. Furthermore, certain tracks are significantly longer than others and certain artists are more likely to create longer tracks than others, so the playcounts of certain tracks, albums, and artists are more likely to have higher playcounts than others. Additionally, keeping track of playcounts still doesn't solve the problem of users not having all the music they know on their playlist, so still part of their range of musical comprehension is overlooked.

Ratings

Methodology

Pros and Cons