| dc.description.abstract | The prominent success of music streaming services has brought increasingly complex challenges for music recommendation. In particular, in a streaming setting, songs are consumed sequentially within a listening session, which should cater not only for the users historical preferences, but also for eventual preference drifts, triggered by a sudden change in the users context. In this dissertation, we propose a novel online learning-to-rankapproachformusicrecommendation, aimedtocontinuouslylearnfrom the users listening feedback. In contrast to existing online learning approaches for music recommendation, we leverage implicit feedback as the only signal of the users preference at each point in time. In a space of millions of songs, we represent each song in a lower dimensional space of continuous features. Our thorough evaluation using listening sessions from Last.fm demonstrates the eectiveness of our approach at learning faster and better compared to state-of-the-art online learning approaches. | |
