Advancing Music Streaming Personalization Developing a Mood-Aware Playlist Generator

In recent years, music streaming services have transformed the way audiences experience and interact with music, with platforms like Spotify leading the charge. As of 2024, Spotify's user base has continued to expand, reaching millions of listeners globally who have access to an extensive library of songs, playlists and podcasts. This shift toward digital music consumption has been accompanied by an increased demand for personalization, where users expect curated experiences that resonate with their individual tastes, moods and preferences. The motivation behind this research is to enhance Spotify’s user experience by creating a system that customizes playlists based on the listener’s current mood and preferences. While recommendation algorithms are widely implemented in various streaming platforms, many still rely on basic filtering methods, focusing primarily on genre or artist. This thesis aims to address a gap by introducing a virtual assistant capable of dynamically generating playlists that reflect both emotional states and musical preferences. Such a tool not only enriches the music streaming experience but also aligns with the broader industry trend toward personalized digital interactions. This study presents the “Spotify-Feel the Music” playlist generator, an innovative approach designed to enhance user engagement through intelligent playlist suggestions. By leveraging the Spotify API, the generator draws on a combination of user mood and historical preferences, providing an interactive experience, both personalized and adaptive. This system underscores the importance of AI-driven recommendations and virtual assistants in creating meaningful, user-centric experiences within digital platforms.