A large number of music resources are provided in the world-wide scope. However, users have to listen to each music data to access and choose their favourite ones, since conventional music databases have not concerned fast browsing and easy comparison of entire search results. In such situation, music collections require smart methods of searching and browsing. In this paper, we propose a novel music database system equipped with a musical knowledge base that assists database processing automatically by providing formal intelligence about music. The musical knowledge base sets appropriate parameters for database processing. It offers an easy and intuitive mechanism for general users to retrieve musical compositions. One of the most significant advantages of our database system is that it introduces a metadata model dealing with sensibility of end users. Our approach makes it possible to realize intuitive music discovering. We show several experimental results to clarify the feasibility of our method.