A Survey of Music Recommendation Systems with a Proposed Music Recommendation System

Abstract

With the advent of digital music and music-streaming platforms, the amount of music available for selection is now greater than ever. Sorting through all this music is impossible for anyone. Music recommendation systems reduce human effort by automatically recommending music based on genre, artist, instrument, and user reviews. Although music recommendation systems are widely used commercially, there does not exist any perfect recommendation system that can provide best music recommendation to the user with the minimal user effort. In this paper, we reviewed the various recommendation systems that are currently in use including content-based, collaborative, emotion-based, and other techniques. We have also explored the strengths and weaknesses of each recommendation technique and at the end, we have provided an overview of a music recommendation system that may solve many of the challenges that existing recommendation systems face through an improved hybrid recommendation system.