Weights Estimation in the Completeness Measurement of Bibliographic Metadata

Abstract

The weighted completeness metrics for metadata use a weighting factor to indicate the importance of each field. In the case of bibliographic metadata, a common way of representing the importance of a field is its frequency of appearance in a given repository. The inaccuracy of this method is why we need to recalculate the weights as the volume of the repository grows. In this paper, we used the Particle Swarm Optimization (PSO) method in the estimation of the weights for the completeness metrics of bibliographic metadata. This method is independent of the metadata format, of the collection and the volume of the repository used. As part of this work, we defined the fitness function of the PSO method to reflect the importance levels of the fields. Finally, we presented a case study with the estimated weights and the calculated completeness of the bibliographic records described at the full cataloging level in MARC 21 format.