A Generic Quality-Focused Data Management Process that distinguishes between semi-structured Research Data, Data Models and Data Transformations

Research data, associated data models and data transformations are often subject to continuous change and high-quality expectations. Hence, ensuring quality is a particular challenge in research and quality assurance is key. Data management processes are known and increasingly used in practice, but, like the ISO Standard 9001 and the GFBio Life Cycle [RFII 2019], they handle Quality Assessment as independent tasks and activities. As a result, data quality is not continuously ensured in data management processes since data quality is not the focus. Furthermore, data quality is closely related to data model quality and data transformation quality, and data quality management processes implicitly address data models and transformations at best. This can indirectly reduce data quality, as poor quality of data transformations and models directly impacts data quality. Therefore, this document defines a quality-focused data management process (QDMP) that distinguishes between semi-structured research data, data models, and data transformations. A meta model fully structures and complements the process with activities, tasks, techniques, artifacts, capabilities, and roles. These key concepts indicate the interrelationships and dependencies of activities and, thus, between data, data models, and data transformations. The process results from a three-year research called KONDA, focusing on the quality of research data