Measuring data quality with weighted metrics

Data quality (DQ) has been defined as 'fitness for use'. In order to measure and improve DQ, various methodologies have been defined. A DQ methodology is a set of guidelines and techniques that define a rational process to measure and improve the quality of data. In order to make DQ measurement and improvement more organised, DQ dimensions have been defined. A dimension is a single aspect of DQ, such as accuracy, completeness, timeliness, and relevancy. In order to measure dimensions, special tools have been developed. These are called metrics. In most organisations, some data are more significant than others. In other words, some data carry more 'weight'. Hence, they must play a more important role in DQ measurement. Most metrics developed so far do not take into account data weights. In this paper, new metrics based on data weights are defined in order to make them more practical. The effectiveness of the new 'weighted metrics' is tested in a case study. The case study shows that the DQ measurements by weighted metrics more closely reflect the opinion of data users.