Emily Tseng wanted to marry Spotify's rich trove of data and easy-to-use API with real-time data visualization to add some live insights to her music. So she made an app to do just that.
Moody Tunes uses Google's natural language processing API to analyze the sentiment of each lyrical line in a user's current Spotify track. The app plots sentiment data over progression through the track. With this, users can visualize the emotional arc of their songs.
Plotting song lyrics line by line reveals some interesting insights, for example the difficulty natural language parsers have in understanding contextual nuance. Future directions for this project include adding visualizations for Spotify's audio data, like danceability, key, and mood, and other NLP analyses, like tone and semantic syntax.
Data pulled through the Spotify and Google Cloud Natural Language Processing APIs. Created with React, Redux and Victory.js, an open-source D3 wrapper designed to work seamlessly with React.
Project Members: Emily Tseng