关键词: 科研数据政策, 数据生命周期, 利益相关者, 机器可操作, FAIR原则

Abstract: [Purpose/Significance] Analyzing the orientation and characteristics of international research data policy can provide an important basis for improving research data policy guarantee system and optimizing research data management and sharing work in China. [Method/Process] This article summarized the two orientations of international research data policy which named as "Integrated DLC-SH" and "Machine Actionable" by using the methods of case analysis and literature research, and analyzed and verified the similarities and differences between the two orientations by using the method of model sample. [Result/Conclusion] It is found that the orientation of "Integrated DLC-SH" adheres to the overall view of research data management and sharing, follows the basic framework of "organization management - business process", and aims to build a global collaborative data governance general plan. The orientation of "Machine Actionable" provides some guiding principles of sequential progression, and attaches importance to the auxiliary role of the data infrastructure, which is in order to realize the global layout of "Internet of Data". The two orientations are different in theoretical basis, core components and service objects, but there are basically the same in basic environment, implementation approaches and final goals, and the integration of the two orientations has become the mainstream trend at this stage. The useful enlightenment to China is drawn up, including: strengthening the guiding function of policy and integrating into the international working system, reshaping the working mechanism and process and improving the data governance pattern, and "combining humanities with technology" to build a "man-machine cooperation" ecology.

Key words: research data policy, data life cycle, stakeholder, machine actionable, FAIR principles