A JXTA-based music information retrieval system
ISSN der Zeitschrift
Information Systems Technology and its Applications, 5th International Conference ISTA 2006
Regular Research Papers
Gesellschaft für Informatik e.V.
In this paper, we present a JXTA-based system for contents-based music information retrieval. The system finds matching melodies from a set of XML documents that encode music contents. The XML documents are stored in a native XML database and XPath query language is used to extract the information about the structure of music data. The matching algorithm utilizes the geometric hashing technique [Wo97] by analyzing geometric properties of both XML documents and music contents. The system preprocesses extracted music contents and stores the information about melody patterns in a relational database. When a user sends a query about a melody, it searches melody patterns that have similar geometric properties as the query melody.