A Pattern for Analysing Data Usage Requirements in Data Sharing

Abstract

A fundamental aspect of specifying data usage in the context of business-to-business data sharing is the definition of data usage conditions, technically represented as policies. Policies can be multifaceted and complex and are primarily used in a case dependent way. This often leads to challenges in the requirements engineering process, e.g., unclear requirements, complex representations, and/or inconsistent interpretations of policies. To address these problems, we introduce the Data Usage Policy pattern that defines the syntax of policies in data sharing. The pattern helps with the structured elicitation and analysis of data usage conditions and enables the creation of a consistent system design. We derive six pattern primitives and a policy grammar from scientific literature and existing requirements of four selected dataspace initiatives, that show a high level of technical maturity, and present them along with a three-step process for policy analysis. We demonstrate the applicability of our contributions with an application example.