Abstract: Traditional k-anonymity methods cannot resist homogeneity attack and background knowledge attack. In order to solve this problem, this paper proposes a sensitive attribute diversity micro-aggregation algorithm, the algorithm groups l nearest tuples to cluster center into one group, which has l different sensitive values. It extends this cluster based on satisfying the l-diversity, so the anonymous table l-diversity yield by the algorithm satisfies sensitive attribute l-diversity can resist homogeneity attack and background knowledge attack. Experimental results show that the algorithm can generate anonymity table to satisfy the needs of sensitive attribute diversity, and to ensure the availability of anonymous table.

Key words: privacy protection, microaggregation, k-anonymization, l-diversity

摘要： 针对k-匿名方法无法抵抗同质性攻击和背景知识攻击的问题，提出了实现敏感属性多样性的微聚集算法，该算法把距离类中心最近的敏感属性值不同的l个元组聚为一类，在满足l-多样性的前提下对该类进行扩展。实验结果表明，该算法能够有效地生成满足敏感属性多样性的匿名表。

关键词: 隐私保护, 微聚集, k-匿名, l-多样性