SIGNATURE WORK
CONFERENCE & EXHIBITION 2024

In this project, our objective is to develop a personalized music recommendation system that can significantly improve the user experience by providing users with more autonomy. This system allows users to select specific factors as prompts-including genres, artists, and moods (such as happy, sad, or angry)-to receive custom-tailored music recommendations. To achieve this, we utilize datasets from the Million Song Dataset and the Spotify API, applying machine learning algorithms such as K-means, Logistic Regression, and XG-Boost. These algorithms help us cluster music tracks based on genres and mood, creating a structured database that serves as a reference for personalized recommendation requests. We simulate a recommendation system that facilitates user interaction for obtaining song recommendations aligned with their specified prompts. Furthermore, we examine the challenges faced by this system and propose potential enhancements. By continuously improving our system, we aim to offer a more responsive, accurate, and enjoyable music discovery experience for all users.