Defining Requirements for Machine-actionable Data Management Plans

Abstract

Data Management Plans (DMPs) are free-form text docu ments describing data used and produced in research projects. The workload and bureaucracy often associated with tradi tional DMPs can be reduced when they become machine actionable. However, there is no common definition of what machine-actionable DMPs really are. This hinders the com munication between stakeholders and leads to scepticism, or conversely to exaggerated expectations. This paper aims to clarify what machine-actionable DMPs are and provides examples of how involved stakeholders can benefit from them. It describes an open stakeholder con sultation performed by the RDA DMP Common Standards working group. The main objective was to define the scope of information covered by machine-actionable DMPs and formulate an initial set of requirements for a common data model for machine actionable DMPs. To do this we used methodology known from system and software requirements engineering to collect information on how needs for infor mation of particular stakeholders evolve over phases of the research data lifecycle.