Developing a Work-Identifier Allocation Algorithm to FRBRize KORMARC Records

Abstract

Displaying search results by work requires clustering related works by assigning every Machine-Readable Cataloging (MARC) record a work-identifier. Although the manual assignment of work-identifiers may be unfeasible in some cases due to a high number of records, developing an instance-appropriate algorithm provides a convenient solution by enabling automated allocation. This study attempted to find a method to automatically convert the National Library of Korea (NLK) data to Functional Requirements for Bibliographic Records (FRBR) structure. However, there were some problems with the Korean MARC (KORMARC) records of the NLK. The NLK uses principally an added entry field (7XX) than a main entry field (1XX) and most 7XX fields do not have relator terms, and which makes us not to know the representative author of books when there are several contributors (multiple 7XX fields). Accordingly, this study seeks to maximize accuracy in cases of works with more than two contributors without using a main entry field or authority data by investigating the possibility of shifting from KORMARC to the FRBR. Author(s) retain copyright, but ASIS&T receives an exclusive publication license