Music Recommendation System for Million Song Dataset Challenge

TL;DR

This paper describes a music recommendation system that participated in the Million Song Dataset Challenge, using collaborative filtering and user similarity to predict missing listening history, achieving a MAP@500 score of 0.15037.

Contribution

It presents a memory-based collaborative filtering approach applied to a large-scale dataset for music recommendation.

Findings

Achieved MAP@500 score of 0.15037

Utilized user-based similarity for recommendations

Participated in the Million Song Dataset Challenge

Abstract

In this paper a system that took 8th place in Million Song Dataset challenge is described. Given full listening history for 1 million of users and half of listening history for 110000 users participatints should predict the missing half. The system proposed here uses memory-based collaborative filtering approach and user-based similarity. MAP@500 score of 0.15037 was achieved.