Music Recommendation using Facebook, Spotify data and Million song dataset user similarities
Motivation: The introduction of music streaming services such as Spotify, Pandora and Apple Music has given users the opportunity to listen to music anywhere and anytime they want. This has resulted in growth in the music streaming industry by 1 million users every month. The music streaming service Spotify has 140M active users with 60M paid subscribers, 30M songs and 2M playlists. With the revenue being generated, the streaming services try to provide the best experience to users by recommending music which each user will like. With so much content and different user data, Spotify uses user’s implicit and explicit feedback to perform content- based and collaborative filtering to recommend songs.
Objective: Our objective is to generate a playlist automatically using user’s Facebook music likes and finding similar artists using “Million Song Dataset – by LabROSA and EchoNest” and their top tracks from Spotify.