An Anonymous Algorithm for Hierarchical Clustering Based on K-Prototypes

摘要

By the research on the clustering problem of multiple attributes data processing based on K-Prototypes algorithm, this paper improves distance formula, which can more accurately reflect the differences between tuples. Besides, according to the various demand of privacy preservation, the sensitive value is divided into multiple levels by (KLS, λ-clustering) - hierarchical anonymous model. The experimental results show that this algorithm is able to achieve highly accurate clustering results. It can also satisfy the requirements of multi-level privacy preservation of sensitive attributes, and effectively reduce the information loss.