Welcome to MoodyTunes! This site is aimed to answer a simply worded, but curious question: What songs make people feel a desired emotion? Through analysis of song metadata and analyzing user feedback, we can collect data that enables us to better understand how you engage with the songs you listen to.
You start by browsing for songs for a given emotion, like "Happy" or "Melancholy". We automatically add songs you upvote to your playlist for that emotion, so you can easily view your collection of songs that you believe make you feel a specific emotion. You can optionally provide a context to grant greater insight to the circumstances involved in your music listening session. This can help you understand the difference in emotion given a certain environment (maybe you find certain songs exciting a party, but calm when you're working on something.)
The site works by analyzing metadata of song's emotional affect. Currently, we analyze the valence, energy, and danceability of tracks. We use the Spotify API to gather this information for use in our site operation. Roughly speaking valence denotes the mood of the song, energy marks the intensity of the song, and danceability indicates how likely the song is to make you move. As you vote on songs for a particular emotion, we use that information to suggest more relevant songs for you as you browse songs for that particular emotion.
Through the course of using the site, we'll track your engagement with the songs you vote on. We'll track your votes for songs you browse (whether or not you vote them as making you feel an emotion) so we can suggest songs that better match your preference for an emotion. If you have any questions about how your data is collected or want to have us purge your information from our records, you can email us your request to our webmaster: ndibari@fordham.edu.