数据治理的场景识别与沙箱机制研究：人机交互视角下的理论框架

Context Identification and Sandbox Mechanisms in Data Governance: A Theoretical Framework from a Human-computer Interaction Perspective

摘要:

[1.1][ly1.2]本文基于人机交互理论视角，构建起“价值—安全”数据治理场景矩阵，进而结合“高价值—高风险”数据治理场景，提出了“建模（Modeling）、实验（Experiment）、诊断（Diagnosis）和解释（Interpretation）”（简称MEDI）四阶段数据治理沙箱机制。研究表明：①人机交互理论为数字经济时代的数据治理提供了全新视角和理论方法，数据治理并非先安全合规、后开发利用或先开发利用、后安全合规的线性过程，而是需要实现数据开发利用和安全合规的动态平衡。②基于价值密度和安全风险两个维度，可以构建起“价值—安全”数据治理场景矩阵。数据治理的重点和难点在于“高价值—高风险”场景，即数据开发利用价值大但未获有效保护或被非法利用损失也大的场域。③数据治理MEDI沙箱机制可以通过人机交互实现规则固化、问题溯源和规则再生，推动沙箱由技术隔离工具升级为可验证、可追溯、可再生的动态数据治理系统。上述研究成果为企业落实《数据安全法》第十三条“以数据开发利用促进数据安全”“以数据安全保障数据开发利用”的制度安排，提供了可嵌入、可复用的数据治理机制框架。

Abstract:

A value–security data governance scenario matrix is constructed from the perspective of human–computer interaction theory. A four-stage data governance sandbox mechanism, including Modeling (M), Experiment (E), Diagnosis (D), and Interpretation (I), is further proposed for the high-value–high-risk data governance scenario. The following findings are obtained. First, a new theoretical perspective and analytical approach are provided for data governance in the digital economy by human–computer interaction theory. Data governance is not regarded as a linear process in which security compliance precedes data utilization or data utilization precedes security compliance. A dynamic balance between data utilization and security compliance is required. Second, a value–security data governance scenario matrix can be established based on the two dimensions of value density and security risk. The key challenge of data governance is identified in the high-value–high-risk scenario, where substantial value can be created through data utilization, while significant losses may occur if data are not effectively protected or are illegally exploited. Third, the MEDI data governance sandbox mechanism enables rule solidification, problem tracing, and rule regeneration through human–computer interaction. The sandbox is therefore transformed from a technical isolation tool into a dynamic data governance system characterized by verifiability, traceability, and regenerability. An embeddable and reusable data governance mechanism framework is provided for the implementation of Article 13 of the Data Security Law, which requires data security to be promoted through data utilization and data utilization to be safeguarded through data security.