关键词: 模糊集, 隐私保护, 隶属度, 模糊偏移度

Abstract: Data transformation is an important approach to data privacy preserving.This paper concentrates on the issue of maintaining the usability of data after transformation and proposes a method based on fuzzy sets.The method transforms sensitive attribute values into fuzzy values and publicizes the data with fuzzy offset degree.This helps the end user to tell the different between two attribute values，even though they are mapped to the same linguistic term.The experimental analysis is designed by adopting clustering algorithm k-means on primitive datasets and perturbed ones by fuzzy sets.It demonstrates that the method efficiently preserves privacy information and maintains the clustering model of primitive data well.

Key words: fuzzy sets, privacy preserving, membership degree, fuzzy offset degree