/PlaylistPulser

Create playlists that peak in intensity at desired times, alternate in popularity, etc. For optimizing crowd energy expenditure and engagement.

Primary LanguageJupyter Notebook

Playlist Pulser

This is an in-development project designed to allow Spotify users to craft the perfect playlist for any occasion. For example, if you want your Halloween party playlist, which starts at 6:30pm, to have a more chill vibe while people arrive but then pulse in energy to a peak in intensity and danceability from 10:00pm - 1:00am, you can do that. If you also want said playlist so alternate between crowd favorites and your personal favorites, you can also do that while maintaining the pulse trajectory.

The code relies on the Spotipy package, which is a wrapper for the Spotify API. You can use it to extract features from music on Spotify, like danceability, energy, valence, popularity, and acousticness. This package provides algorithms that use these features to generate playlists depending on user preference.

Future implementations will integrate multisensor audiovisual streams to create an adaptive party environment, where the algorithm pays attention to the crowd's collective social state and optimizes energy expenditure and enjoyment.