An open source platform to assist the creation of group playlists through artificial intelligence algorithms

Recommendation systems are a constantly expanding study area, with applications in various fields such as e-commerce, films, music to promote the user’s suggestions. When we talk about music, we have more than 20 years of studies trying to solve the problem of a good generation of playlists that maximizes the satisfaction of a larger number of listeners. For automated automatic playlist generation methods focusing on a user group, we have the collaborative filter as a more assertive method to get the user’s not likely, to improve the performance of group recommendation algorithms we store the preferences of users Especially I did not like it by placing the availability of using this data as an algorithm input parameter. The platform described in This paper is intended to facilitate testing between these recommendation systems, standardizing data entry, and facilitating requests. The use of GraphQL as a framework associated with Apollo as a library, greatly facilitates the integration of these APIs, as the separation of data sources makes it possible to associate Spotify data with Deezer or Apple Music data, these data are stored in the database of the connection, so that in future requests it will no longer be necessary to consult the Spotify API, thus facilitating the consumption of data from the artificial intelligence algorithms, as well as a possible sharing of songs between services, since all services have an ISRC code to identify the songs.