Back in May of this year I posted a link to an article comparing various music recommendation engines, with Last.fm and Pandora coming out on top of the heap. Now, Steve Krause shares his thoughts on the differences between the different approaches taken by Pandora vs Last.fm and which works best. Steve frames his thoughts with the nature vs nurture debate:
Taking the nature side, Pandora’s recommendations are based on the inherent qualities of the music. Give Pandora an artist or song, and it will find similar music in terms of melody, harmony, lyrics, orchestration, vocal character and so on. Pandora likes to call these musical attributes “genes” and its database of songs, classified against hundreds of such attributes, the “Music Genome Project.”
On the nurture side (as in, it’s all about the people around you), Last.fm is a social recommender. It knows little about songs’ inherent qualities. It just assumes that if you and a group of other people enjoy many of the same artists, you will probably enjoy other artists popular with that group.
Steve tackles a number of attributes of each systems approach in his article. For instance, Pandora employs a staff of musicologists who listen to music and apply their experience and expertise to identify attributes that allow the “qualities” of a song to be identified and tagged, leading to the ability for the Pandora system to make recommendations accordingly. The big advantage to this approach is that new music/artists stand a better shot of more immediate recommendation. Last.fm uses social collaborative filtering to build its database, which is arguably more accurate over time but can make it difficult for new music to gain a proper foothold early on. Other comparisons are discussed in Steve’s blog post which is a worthwhile read.
Steve makes the point that Pandora has the greater opportunity as they can more easily integrate elements of Last.fm into their system than the other way around, but Pandora as a business has a far higher burn rate due to a far greater overhead. A combined approach would certainly be interesting, but regardless of where it goes I continue to use both and recommend both services very highly to anyone who loves music. A little investment of time in both places will unquestionably give you some good recommendations of new music, or remind you of some lost gems you’d forgotten.