Music Recommendation System Using Hybrid Approach

Artificial Intelligence & Machine Learning

In the digital music era, artists and creators on platforms like Spotify seek to engage and connect with their audiences. To facilitate this interaction and promote their playlists, a novel hybrid-based recommendation system is proposed. This project presents a hybrid music recommendation system that combines content-based and collaborative filtering methods to offer personalized song recommendations to users. Leveraging the Spotify API and a rich dataset of song features, this system offers an interactive and user-centric experience for music enthusiasts. The content- based filtering component analyzes audio features of songs, including danceability, energy, acousticness, instrumentals, and more. Users can input their favorite songs, and the system generates recommendations based on the similarity of these features. Additionally, users have the flexibility to fine-tune their preferences using sliders for various audio attributes. The collaborative filtering component employs K-Means clustering to group songs with similar audio characteristics. When users provide input songs, the system calculates a user profile based on the mean audio feature values of those songs. Recommendations are then generated from the cluster that best matches this profile, ensuring contextual relevance. Integration with the Spotify API enriches the dataset with song popularity, explicitness, and additional audio features. Users can listen to recommended songs directly from the interface, enhancing their music discovery experience. This hybrid recommendation system provides a dynamic and adaptable approach to music discovery, catering to diverse user preferences. Its seamless integration of content-based and collaborative filtering techniques empowers users to explore new music while staying connected with their musical tastes.

This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.