Ethnic music exploration guided by personalized recommendations: system design and evaluation

Abstract

Ethnic music collections provide vantage points into the culture of different communities. However, tools that allow the exploration of those music libraries by the general public are lacking. We present and evaluate an ethnic music collection exploration system that comprises a recommendation algorithm and a web-radio interface. The proposed system provides recommendations that adapt to the user's preferences. A blind a/b user study conducted to evaluate the system suggests that personalized recommendations could help individuals from diverse backgrounds explore ethnic music collections. We believe both the system and its evaluation contribute to the growing work in the field of intelligent information systems.