Infoscience

Report

User Acceptance Issues in Music Recommender Systems

Two music recommender systems were compared side-by-side in an in- depth between-subject lab study. The main objectives were to investigate users' acceptance of music recommendations and to probe the main technology acceptance model in the environment of low involvement recommendations. Our results show that perceived usefulness (quality) and perceived ease of use (effort) are the key dimensions which are sufficient to incite users to accept recommendations, and that the adapted model is suitable for entertainment recommenders. Measures of quality such as accuracy, enjoyment, satisfaction and having music tailored to a user's taste are directly correlated with acceptance, and measures of effort like the initial time to reach interesting recommendations and the ease of use for discovering music are strongly linked to acceptance. The study shows how important it is for a music recommender system to take into account users' emotions and mood. Finally, the results highlight the necessity for low- involvement recommenders to be highly reactive.

Related material