The primary goal of project SongBird is to develop a music-recommendation engine that will predict songs for a user based on songs they have listened to previously. The user should be able to interact with this iteration of the engine through a web-interface. This can involve a light version of the engine that runs stand-alone in a web-browser or a heavier version that serves predictions from a remote server.
The secondary goal of the project will be to package the engine in a lean format that can efficiently run on embedded hardware. This will then be handed over to MDB to possible embed into an iOS/Android mobile application.
Songbird Demo DD1: https://ml.berkeley.edu/dd/songbirddemo.html