Spotify playlist generator using sentiment analysis on personal journal/diary entry data. Requires Spotify Premium to use; I've hidden my user credentials in but if you'd like to run this project, download the below dependencies and visit https://developer.spotify.com/ to generate your personal tokens. See the "Example" section below for a demonstration.
- Spotipy
- Access Spotify API for user history and playlist generation.
- NLTK
- Valence Aware Dictionary and sEntiment Reasoner (VADER) for sentiment analysis of user text entry.
- Pandas
- Storage of scraped song data and Boolean indexing for quick recall.
- sentiment.py
- Evaluates the sentiment of a given body of text.
- spotify_scraper.py
- Scrapes Spotify API for song information and contains details in a pd.DataFrame.
- audio_thera.py
- main() function that runs scripts developed in the other .py files.
- secrets.py (.gitignore)
- Contains user credentials including username, scope, client_id, client_secret, and redirect_uri for the app.
TODO: Add screen recording of script in action (include arg from command line, Spotify API auth, and resulting playlist)
- Host application on Flask or Django.
- Create SQLite database to store sentiment analysis results / recommended songs.
- Improve method of analyzing song sentiment through Spotify's metrics.
TODO: Add MIT license data here.