Collective Music

Collective Music, was a demo of a semi-private community, where I tried to get a large shared music collection, where the community helped supply the music, edit the album information, and recommend good albums to each other. The main purpose of the site was to collect user interactions with the music, to be able to recommend new music using a collaborative learning algorithm.

Collaborative Learning Algorithm
The algorithm was to group users into similar listening habits, so that the system could later recommend new albums to users based on what peers with similar tastes are listening to. Grouping of users was not based on details of the albums, but on the patterns of listening. Music was not classified into genres; by ignoring the genre, I hoped to be able to make recommendations that were not affected by incorrectly set genres or other classifications. For more details check out my report on it here.

Features
Collective Music had a large feature set including: user referals, custom avatars, chatting, online player, custom playlists, download single song, download complete album, upload songs, recommend songs to the community, and much more.

Demo
Here is a demo version of the project. The demo does not play any music, or allow you to download any music.

Username: demo
Password: demo123
Demo: Click Here

Screen Shots