Data is the cornerstone of critical business processes in most organisations. In this context, organisations are responsible for determining the characteristics that data must follow to decide whether or not to use it. The usability of data strongly depends on the context in which the data is to be employed, and on the context in which the data is generated.Bearing this in mind, we defined a methodology (DMN4DQ) to automatically generate recommendations on the usability of the data supported by a decision process, including a context-dependent data-quality assessment based on Business Rules for Data Decisions, modelled in a hierarchical structure. In order to support the decision process, and to enable data quality stewards to define their decision rules, we rely on Decision Model and Notation (DMN). This standard provides a declarative mechanism to define a decision logic model that is understandable by non-expert users.Our proposal is validated in a case study. We developed a tool (dmn4spark) to apply the decision logic to the dataset. After the execution, data stewards can filter non-usable data, reducing the risks associated with bad-quality data in their business process, and identifying the root cause of non-usable records.The usability of data depends on the data quality and on the context in which the data is used. Our methodology provides a hierarchy to integrate decision rules about (i) data values+ADs (ii) measurements of data quality dimensions+ADs (iii) assessment through the aggregation of dimensions, and (iv) data usability. The use of DMN makes the transformation of knowledge into a formal model easier, facilitating the automation of the generation of recommendations.