Construction of a Theoretical Framework for Scientific Data Governance

The advancement of data-intensive sciences and artificial intelligence-driven sciences has introduced governance challenges for multi-source heterogeneous scientific data across diverse scenarios. Given the intricate entanglement of stakeholders, processes, and content in scientific data governance, this study intends to propose a theoretical framework to elucidate its complex dynamics and inform governance practices. The theoretical framework for scientific data governance consists of three core dimensions: data stakeholders, data lifecycle, and data governance elements. Non-systematic literature review was employed to identify the classification of data stakeholders and data lifecycle, and bibliometric analysis was used to extract the elements of scientific data governance. Meanwhile, based on the elements of data governance, five governance systems have been summarized, including organizational operation system, technical support system, risk prevention and control system, value realization system, and regulatory system.