I made this project to help me find new music. As much as I enjoy using Spotify I find that the music it recommends to me is mainly from arists that I've already heard. For this project I used Spotify's "hidden" features ('energy', 'danceability', 'speechiness', 'acousticness', 'instrumentalness', 'liveness' and 'valence') and found the cosine-theta simlarity between the features for a target track and a dataset of around 600k songs. The top twenty songs with the highest similarity are displayed.
I created a basic website using django which is hosted here.
To run this project locally download the dataset from kaggle and place the "tracks.csv" file in the main directory. Dataset is from https://www.kaggle.com/yamaerenay/spotify-dataset-19212020-160k-tracks. Then run "python manage.py populate_db tracks.csv". This will populate the database with rows from the csv file.
Background image taken from pexels.com, link to original https://www.pexels.com/photo/people-enjoying-the-concert-1047442/