Preenchimento de Playlists Utilizando Técnicas de Sistemas de Recomendação Baseadas em Filtragem Colaborativa

Abstract

With the rapid expansion of digital music, manage and search for songs has became significant. With this expansion, the music recommender systems plays a fundamental role in this process. But due to subjectivity of music and its listeners, it is not possible do determine with total precision which music to recommend. This way playlists, list of songs, can be more precise to the user’s need, as to their mood and moment. Currently a recommendation approach that has been proved efficient is the collaborative filtering, where the data from different users is used to build recommendations. This paper presents a music recommender system to fill playlists using collaborative filtering, based on songs popularity and similarity between playlists.