Extending the Personal Music Collection by Free Music

Today, people listen to music on their mobile phone or other playback devices. Jukefox is a special music player running on Google’s smart phone operating system Android. It’s capable of recognizing similarities between multiple songs, albums and artists using data based on a huge music similarity map [1] which classifies songs into a high-dimensional space. Songs with small distance are similar, while widely separated coordinates mean, that two songs are very different. This thesis shows how the service of jukefox can be extended with the functionality to explore new music that is similar to the one the user already has in his own music collection. It describes, how artists and songs that fit into the current mood and the overall music taste of a user, can be found based on the already available music similarity map and how a user can listen to them without the need of downloading every single piece of music manually. As sources for new music, services like Jamendo5, the Free Music Archive6, ccMixter7 or SoundCloud8 with music licensed under the Creative Commons perfectly qualify. In this thesis a proof-of-concept implementation for Jamendo as music source is introduced, while the architecture and design allows for simple extension by other sources.