Goombah
Goombah is an intriguing new application/service that examines your iTunes music library and compares it with other users’ libraries. It then recommends tracks that you might like and lets you listen to them using the iTunes Music Store.
- You should know that other Goombah users can browse the entire contents of your music library, including play counts and ratings (but not comments).
- It’s not clear to me how it knows which of the music in my library I like. For instance, two of the top four tracks it recommended to me were Carmen and Wagner, although I have not rated or played any classical music since starting my current iTunes library.
- Most of the recommended tracks are not (yet) available at the iTunes Music Store.
- It would be great if a future version could let me rate the recommended items, as Amazon does.
- It’s written using PyObjC.
4 Comments RSS · Twitter
Hi Michael,
In this first release we pay no attention to ratings. Instead, we use user's frequency of playing each song.
Later we will pay attention to ratings entered through the normal iTunes user interface.
Note we're fine tuning the recs over the next day or two... don't take them seriously yet... we needed some real data to plug in and now we have it.
Thanks for posting about us!! :)
Cool. I'm excited. This looks like the half of Amazon that ITMS has been missing, and it's potentially much more accurate and interesting.
Will it hurt my recommendations that I've edited a lot of my ID3 tags so that they are no longer exactly as they appear in CDDB?
Glad to see Gary and crew kept with it!
Robb "I read the Emergent Music whitepaper back in the day" Beal
Reminds me a bit of AudioScrobbler. I've contemplated trying to add support for some sort of collaborative filtering service to PodWorks, but so far I've never really found one I liked.
