This paper provides an overview of four years of the authors’ research on music information retrieval (MIR), omitting technical details. An overview of terminology, music analysis techniques, and research done in the last five years is also provided. Some of the results achieved by the authors are summarized, for example distinguishing among several genres of music by automatic classification. However, the complete lack of detail in the discussion of MIR research limits the possibility of a deeper understanding of this subject. The paper continues with a section devoted to graphical user interfaces designed for browsing and interacting with audio collections. Last, the authors’ own system, MARSYAS, is introduced.
Illustrative of the phase of development that MIR is in is the statement that a database of 1,000 files is large; in the field of information retrieval, and even for content-based image and video retrieval, such a size is considered small. This illustrates that research toward MIR is still in its early years. Regrettably, no references are provided for the last three years. Nevertheless, for nonexperts, this paper provides a nice and easy introduction to the field of MIR, offering some pointers for further reading, and providing a link to public license software.