Skip to content

Develop a system that personalizes music recommendations for users and analyzes their engagement and retention.

Notifications You must be signed in to change notification settings

scottpitcher/spotify-user-engagement

Repository files navigation

🎵 Spotify Playlist Recommender and User Engagement/Retention Analysis 📈

🎯 Objectives

  • Build a recommendation system to suggest songs based on user preferences
  • Predict user engagement levels
  • Analyze factors influencing user retention and predict churn

📊 Data

  1. Songs Data: Song ID, Title, Artist, Album, Duration (s), Popularity, Release Date

Source: Spotify Web API and connecting through spotipy library in Python

  1. User-Engagement Data: User ID, Song ID, Play Count, Last Played, User Age, User Country

Source: Synthetic data generated from Spotify API data*

*For this project, the Spotify API'd data will come from a selection of public, Spotify-created playlists; these playlists will be both Global and Country-specific. Country-specific data will be used to create more accurate synthetic user-data from those respective countries, while Global songs will be scattered through all users listening histories. The code for this process can be found in user_data_generation.py script