Music Information Retrieval (originally posted 1-24-2009)

The relatively new domain of Music Information Retrieval or MIR is a rapidly evolving, technology-driven recent entrant on the information retrieval scene. Generated by information scientists, computer scientists, engineers, mathematicians, and musicologists, among others, the domain has contributed new systems for automatic storage and retrieval of music. Mapping the domain is itself a fascinating business. Recently I asked “Music Information Retrieval: An Example of Bates’ Substrate?” in a paper for the Canadian Association for Information Science/L’Association canadienne des sciences de l’information. This is the abstract:

Bates suggested that the intrinsic unity of information science lies in ‘substrate’-the properties of information and its transmission. Music Information Retrieval (MIR), and ISMIR annual conferences offer a rich panoply of intellectual and cultural diversity. We map the evolution of MIR using conference papers from 2000 through 2005. Results indicate tight thematic coherence in the domain around the problems of information retrieval and classification, and the locus of most research within computer science departments.

The paper is available here: http://www.cais-acsi.ca/search.asp?year=2006.

Author co-citation analysis was also revealing: indicat[ing] tight thematic coherence in the domain around the problems of information retrieval and classification, and the locus of most research within computer science departments. Citation practice indicates the habits of a hard science. Author co-citation within the domain is abundant, J. Stephen Downie is clearly the founding focal point, but the domain is very focused, reinforcing the notion of a tightly-packed, emerging and continuously successful domain. ACA data from outside the domain provides an interesting comparison; watch for another paper soon.

Advertisements
%d bloggers like this: