Documentation, Informaiton & Knowledge ›› 2025, Vol. 42 ›› Issue (5): 19-30.doi: 10.13366/j.dik.2025.05.019

• Special Topic(1): Trusted Data Space • Previous Articles     Next Articles

The Construction and Long-Term Operation Path of High-Quality Dataset Under Collaborative Governance in Trusted Data Spaces

LIN Zhenyang1,2, HU Xin3, GUO Mingjun4, HE Yue5, ZHANG Yan6, WANG Mingzhu7   

  1. 1. Multimodal Artificial Intelligence Data Engineering R&D Center, Tsinghua University Shenzhen Graduate School, Shenzhen, 518071;
    2. Wuhan Institute of Data Intelligence, Wuhan, 430000;
    3. School of International Politics and Economics, University of Chinese Academy of Social Sciences, Beijing, 102401;
    4. Big Data Development Department, National Information Center, Beijing, 100045;
    5. Renmin Law School, Renmin University of China, Beijing, 100872;
    6. Peking University Law School, Beijing, 100871;
    7. School of Management, Capital Normal University, Beijing, 100048
  • Online:2025-09-10 Published:2025-11-13
  • Contact: Correspondence should be addressed to GUO Mingjun, Email: guomjnature@126.com, ORCID: 0000-0002-6767-2448
  • Supported by:
    This is an outcome of the Special Project "Research on the Functional Positioning, Operational Mechanism and Governance Mechanism of Data Trading Venues"(72442030)supported by National Natural Science Foundation of China, and the 2024 Research Project "Research on the Construction of High-Quality Data Sets in Hubei Province"(IM2409E085N1)supported by Hubei Provincial Data Bureau.

Abstract: [Purpose/Significance] Aiming at the challenges of quality inconsistencies, circulation barriers, and unbalanced rights and responsibilities in the integration of data elements between government and enterprise, guided by the needs for high-quality dataset construction and operation in the context of digital-intelligent integration, this study explores a collaborative governance pathway driven by trusted data spaces, with the aim of resolving multiple dilemmas currently faced by datasets, such as limited quantity, poor quality, and difficulty in use. [Design/Methodology] By constructing a four-in-one integrated operation platform architecture for high-quality datasets based on trusted data spaces, encompassing "data quality improvement–data marketplace–data computation integration–data crowdsourcing innovation," this study designs a long-term operation mechanism with three-level linkage of "city-industry-enterprise". It analyzes the governance rules, technological adaptation, and scenario coupling logic in the integration of government and enterprise data elements driven by trusted data spaces. It proposes a three-pronged promotion paradigm of "scenario driven-mechanism coordination-security guaranteed". [Findings/Conclusion] The construction of the high-quality dataset integrated operation platform achieves the efficient circulation of the entire dataset process,value realization, and a virtuous cycle of data feedback. The development of a data quality "feedback" model forms a closed-loop mechanism of "circulation–value addition–quality improvement", validating the effectiveness of the collaborative governance framework of "regulatory constraints+technical support+commercial incentives". [Originality/Value] This study puts forward the theoretical expansion and practical pathways of "four-in-one" high-quality dataset construction model based on trusted data spaces, breaking through the limitations of the traditional model that "emphasizes construction over operation and technology over institutions". It provides references and decision support for the industry to implement the action plans of "data elements ×" and "artificial intelligence +" in China.

Keywords: Data spaces, Data elements, Government-enterprise data, High-quality datasets, Data governance, Digital-intelligence convergence