Repository hosting codebase and Poster for CS 6242 Course Project at Georgia Tech.
Our project was directed towards building a music recommendation product with key focus on new song discovery and explainability. The system is an attempt to generate a playlist by suggesting songs heard by users having similar preferences. The product takes Spotify user-id of the user as an input and generates a music recommendation playlist. The Tableau UI also has toggles to select a mood to update the recommended playlist and has an explanation window for why a song was recommended. We used a hybrid approach - a combination of collaborative filtering and content-based filtering- to develop our recommendation system. The user data will be procured from Spotify for Developers’ Web API and the user-song data is taken from subset of MLHD[6] dataset. The Spotify API provides access to multi-level data such as user’s playlist data, song’s attribute data, artist data etc and the MLHD data has user and song interaction data.
