A FAIR Perspective on Data-Quality Frameworks

Despite considerable effort and analysis over the last two-to-three decades, no single encompassing solution has been found for data-quality frameworks. Currently, the choice is between a number of frameworks dependent upon the type and use of data and they are generally prescriptive of specific quality dimensions. We re-examine the need for measuring data quality with reference to the FAIR data principles and propose an integrated framework that scales over different data types. Our approach builds on an existing federated data-description framework that handles the FAIR-related quality dimensions in the general data contextualisation attributes and describes the other quality dimensions in dedicated associated contextual spaces. The framework provides several advantages – it can handle the quality descriptions at any level of granularity for any data type; it does not blur the quality dimensions between the data and the data-application perspectives; it allows traceability through a chain of data-processing operations providing data-quality provenance; and it is flexible to the extent that any data metric (even at a subjective level) can be specified. Examples have been provided showing how the framework can be used practically and in a following study the framework will be used to describe the data quality of a real-world indicator.