Abstract: In view of the problem of privacy risk situation assessment, a set pair analysis assessment system is set up for privacy risk indicators based on the theory of five-element partial connection numbers. The weights of assessment indicators are determined by the method of least squares partial weighting. Partial differential processes are carried out on different stages by using the theory of five-element connection numbers. Potential values and trends are calculated and analyzed for partial connection numbers on different stages in order to realize situation assessment of privacy risk. The proposed method can dynamically reflect privacy risk assessment indicators and effectively eliminate interference and adverse effects caused by uncertainty, vagueness and randomness during the process of privacy risk assessment. Finally, some examples are given to analyze the feasibility and effectiveness of the proposed method.

Key words: privacy risk, situation assessment, five-element partial connection numbers, least squares partial weighting

摘要： 针对数据隐私保护风险评估问题，利用集对分析的五元偏联系数理论，为隐私风险待评估指标建立了集对分析评估系统。采用最小二偏赋权方法确定评估指标的权重，并运用五元联系数对隐私评估指标进行分阶偏处理，对各阶偏联系数进行势态计算并分析趋势走向，实现对隐私风险态势的评估。该方法能够对隐私风险评估指标进行动态的体现，有效消除了隐私风险评估过程中不确定、模糊、随机等误差因素的干扰和不良影响。通过实例分析了该方法的可行性和有效性。

关键词: 隐私风险, 态势评估, 五元偏联系数, 最小二偏赋权