Music recommendations based on your favorite artists, songs, and listening history powered by machine learning!
By entering your favorite artist and favorite songs from that artist, this program recommends you songs from that artist that are similar to your favorite songs in terms of their audio features (see files DrakeTrackData.csv and DrakeTrackDataClustered as examples).
This recommendation system is done through k-means clustering, which groups each song into a cluster in which all songs in the same cluster have similar audio features. Given that I like songs with similar beats and rhythms, I thought this would be an effective way to find new, catchy music to listen to.
The program also goes through your Spotify listening history and finds other artists and songs you frequently listen to and recommends you similar songs from those artists with the same recommendation system.
-
If you do not have git installed, download it here
-
If you do not have pip installed, download it here
-
Clone the repository
git clone https://github.com/faizancodes/Spotify-ML-Music-Recommendations.git -
Navigate to the downloaded folder
cd Spotify-ML-Music-Recommendations -
Create and activate a virtual environment
-
Read more about VMs here if you are having issues
- Windows:
py -3.6 -m venv env env\Scripts\activate - Mac & Linux
python3.6 -m venv env source env/bin/activate
- Windows:
-
-
Download all the dependencies
pip install -r requirements.txt -
Create a Spotify Developer account here
- Enter your client credentials and spotify username in the
spotifyrec.pyfile
- Enter your client credentials and spotify username in the
-
Run
spotifyrec.pyto execute the code