Basic Study on Automatic Identification of "Works" and "Expressions" for Music MARC : Exploring the Way to Generate Both Entities' Identification Keys

Abstract

Music materials have the features such as multiple pieces in one medium, multiple statements of responsibility, multi-language, plasticity and fragmentation of work, also unsers have search requests associated with these features. However, even for music library OPAC systems at present, it is not able to respond adequately to these features as shown by previous studies. This article describes basic study on automatic identidication of "works" and "expressions" for music MARC as ony way to solve this problem, aiming at "FRBRization" of music library OPAC systems. Specifically, by reference to the results of an analysis of 300 records related to classical music sampled from a music MARC (Toccata MARC), this article explores to way to generate work and expression identification keys. The study ultimately aims at automatic identification of works and expressions for music MARC by clustering identification keys based on string matching. This article is intended to be the first step of the study.