Open datasets for evaluating the interpretation of bibliographic records

Abstract

The transformation of legacy MARC catalogs to FRBR catalogs (FRBRization) is a complex and important challenge for libraries. Although many FRBRization tools have provided experimental validation, it is difficult to evaluate and compare these systems on a fair basis due to a lack of common datasets. This poster presents two public datasets (T42 and BIB-RCAT) intended to support the validation of the FRBRization process.