Music Recommendation Project

The final part of my degree is the final year Masters project. Mine is based on music recommendation, where I'm trying to write my own music recommendation engine.

The approach I'm using is to combine two techniques; user-based collaborative filtering and item-based collaborative filtering, using what is known as similarity fusion. Along with that, I'll be generating implicit ratings based upon certain criteria such as how many times a track has been listened to, how many times a track has been skipped, etc. This is the main bulk of the project, and the part that's causing me the most headaches. But progress is being made, and I'm nearly there.
Read More...