分类分级制度是促进科学数据开放共享和高效流通的规范基础，但现有分类分级制度存在具体实施细则供应不足、正面清单管控模式制约数据流动、分类分级的价值权衡只关注静态安全等问题。为因应开放科学的动态治理需求，提升科学数据的价值释放和应用质效，分类分级制度应由封闭保守的静态安全导向转向统合安全价值的开放共享导向。基于对20个国家级科学数据中心分类分级实践的观察与分析，提出建构开放共享导向型科学数据分类分级制度的建议，具体包括确定公开、合理、无歧视的分类分级立法指导原则，强化政策群协同，落实开放共享绩效考核及问责机制，推动数据正面清单管理制转向负面清单制，建构统合安全和共享价值的科学数据分级判定方法，引入比例原则以支撑科学数据分类分级中的多元价值权衡。

In the modern era, scientific and technological innovation serves as a crucial driving force for economic and social development. In the digital age, research paradigms are leaping towards data-intensive type, with national strategies placing a greater emphasis on the development and utilization of scientific data resources. However, the vast and complex nature of scientific data presents fundamental differences in value, type, and form; scientific research imposes multifaceted requisites on scientific data utilization, encompassing accuracy, sharing, and security. The extant scientific data governance mechanisms struggle to effectively manage the supply-demand dynamics of scientific data, failing to mitigate the negative externalities of data breaches that threaten national security, public interest, and private rights. Moreover, these frameworks are falling short in leveraging the beneficial externalities that could arise from the open and collaborative use of scientific data, which is crucial for driving national progress, bolstering economic growth, and fostering technological advancement.
This paper aims to enhance the scientific data governance system to balance data security with efficient data flow. It highlights the benefits of a data classification and grading system for managing heterogeneous data, aligning with the complex nature of scientific data. However, current research often overlooks the system's role in data circulation and sharing, focusing instead on security. The paper points out a significant gap in the literature regarding the impact of scientific data centers' classification and grading systems on overall governance. To address this, the paper conducts empirical research on China's national scientific data centers, applying contextual integrity theory and policy analysis to assess how data flow supports freedom, security, and value within specific contexts.
This paper utilizes the publicly accessible texts related to classification and grading from 20 national scientific data centers as research samples, and sets up seven observational points, namely normative categories, resource support, policy synergy, management model, liability allocation, openness degree, and normative value, to evaluate the supply, effectiveness, and value balance of the classification and grading system norms of scientific data centers in China. The breadth and depth of the research perspectives are sufficient to ensure the scientificity and reliability of the research results. The results indicate three cardinal issues within the classification and grading system of national scientific data centers. Firstly, the classification and grading system norms of the 20 national scientific data centers are inadequately formulated and have weak cooperativity among norm groups. Secondly, the positive list management and imbalance of liability allocation diminish incentives for the open sharing of scientific data. Thirdly, the existing classification and grading system overemphasizes data security to the ignorance of circulation, failing to achieve an appropriate equilibrium between security protection and open sharing values.
Drawing on the research findings, this paper offers several recommendations for improvement. It advocates for the augmentation of formal regulations that adhere to the principles of openness, rationality, and non-discrimination, while also reinforcing the harmonization of norms across various governance bodies. Subsequently, the paper suggests transitioning from a positive list to a negative list approach, thereby fostering a more permissive atmosphere for the dissemination of scientific data. Additionally, the establishment of an accountability and incentive framework is proposed to encourage the sharing of scientific data. Furthermore, it calls for the refinement of scientific data classification and grading protocols, integrating considerations of security and sharing value, and accounting for the positive and negative externalities associated with data utilization. Lastly, the paper endorses the application of the Principle of Proportionality, encompassing appropriateness, necessity, and balance, in the valuation and categorization of scientific data.
This paper integrates research findings from the domains of Contextual Integrity theory, data classification and grading system, and scientific data governance. It innovatively introduces the Contextual Integrity theory into the fields of scientific data governance and the classification and grading system, filling the empirical gaps in the scientific data classification and grading system. Moreover, it constructs a novel model for classification and grading system model that balances both openness and sharing, and puts forward new approaches to optimizing scientific data governance. The results of this paper are expected to provide valuable guidance for researchers in their data management activities.