基于敏感属性聚类的电子病历隐私保护算法

Privacy Protection Algorithm in Electronic Medical Records Based on Sensitive Attribute Clustering

目的:为了保护电子病历中病人的隐私不被泄漏。方法:在K-匿名算法和L-多样化匿名算法的基础上,不仅仅针对数据关联攻击和同质攻击,还增加了抵抗背景知识的攻击。结果:提出了一种基于敏感属性聚类的电子病历隐私保护算法。结论:所提出的算法比K-匿名算法和L-多样化匿名算法在保护病人隐私方面有更低的隐私泄漏概率。

Objective: Aim to protect the privacy in electronic medical records. Methods: Based on K-anonymous algorithm and L-diversity algorithm, not only to prevent from the data joint attack and homomorphic attack, but also background knowledge attack. Results: A sensitive attribute clustering based privacy protection algorithm is brought. Conclusion: The new algorithm can protect the privacy in electronic medical records better than K-anonymous algorithm and L-diversity algorithm.