SongRecommenderSystem

Introduction

Our project aims to develop a song recommendation system. We analyze a song and evaluate its main features and want to find the most compatible one among other songs and recommend it to the user.

Youtube link

Requirements

• Python 3
• Pandas
• Numpy
• Spotipy
• dotenv

Installation

pip install numpy, pandas, spotipy

# optional package to read from .env file
# client_id and client_secret can be mannually set in the code
pip install python-dotenv

Usage

1. After getting cliend_id and client_secret from Spotify Dashboard, edit data/getdata.py and set them to the related fields.

2. Then edit the playlists as you want in data/playlist_urls.txt file. Each line should contain a playlist url.

3. Then, run the following command to get the track details.

   cd data
   python getdata.py

4. Now data/data.csv file should be created. You can use this file in the notebook.

5. Run notebook.ipynb and enjoy!

References

• Spotify API
• Spotipy docs
• Spotify for Developers

Disclaimer

This project is for educational purposes only. We do not own any of the data used in this project nor commercially use the data. All rights reserved to their respective owners.