Back to Portfolio

UIUC Research Park Hackathon

Prototyped SpotOn, a Spotify music recommendation system that uses the current location and time to suggest songs.

Overview

Our team classified different areas of the Champaign campus into distinct sections, such as the rural farmland south of campus, the bustling and exciting clubs and restaurants on Green Street, and the cozy and quiet middle of campus with many libraries and study spots. We used this information, along with the current time of day for the user, to come up with "feature" scores used by the Spotify API, such as acousticness and danceability. These are used, in addition to the user's data, to determine the next songs played by Spotify's recommendation algorithm.

Technologies

Python TensorFlow Wearable APIs Flask

Key Features