Duplicate detection and record consolidation in large bibliographic databases: the COPAC database experience

Abstract

COPAC is a union catalogue giving access to the online catalogue records of some of the largest academic research libraries in the UK and Ireland. Like all union catalogues, COPAC is supplied with multiple copies of records representing the same document in the contributing library catalogues. To reduce the level of duplication visible to the COPAC user, it has been necessary to develop duplicate detection and record consolidation procedures. These result in the production of a single record for each document, representing the holdings of several libraries. This paper discusses the ways in which both the duplicate detection and record consolidation procedures are carried out, along with the problem areas encountered. The general structure of these procedures is also described, providing a model of the duplicate record handling mechanisms used in COPAC.