This paper presents a semantic model which delivers personalized audio information. The personalization process is automated and decentralized. The metadata which support personalization are separated in two categories: ...