An Analytical Framework to Analyze Dependencies Among Data Quality Dimensions

Dependencies among data quality dimensions, although they are not thoroughly and deeply analyzed in the research area of information quality, represent an issue of primary importance. The knowledge on dependencies can help the data quality analyst in finding errors, it can be useful to assess the quality level of a data set, which can improve knowledge on data quality dimensions. This paper proposes a data-driven approach for the analysis of dependencies among data quality dimensions. The analytical framework proposed here provides the main models of dependencies and analytic formulae, based on the entropy of Shannon. It characterizes each model, and proposes measures of correlation among the dimensions. Automatic identification of models of dependencies for a given data set is performed with the implemented version of the framework. Examples and case study are provided.