Metrics for Measuring Data Quality - Foundations for an Economic Oriented Management of Data Quality

Abstract

The article develops metrics for an economic oriented management of data quality. Two data quality dimensions are focussed: consistency and timeliness. For deriving adequate metrics several requirements are stated (e. g. normalisation, cardinality, adaptivity, interpretability). Then the authors discuss existing approaches for measuring data quality and illustrate their weaknesses. Based upon these considerations, new metrics are developed for the data quality dimensions consistency and timeliness. These metrics are applied in practice and the results are illustrated in the case of a major German mobile services provider.