情報処理学会 第80回全国大会講演要旨

A large amount of data is being collected and published for useful analysis. For example, governments are collecting medical records of individuals to improve health care system. However, sensitive information may be revealed in this process. Removing the person-specific information, such as names and phone numbers, is not enough due to the existence of public data with which someone may correctly identify some person related to the sensitive information. This threat led to the privacy preserving paradigm of k-anonymity. In this paper we design an algorithm to solve a generalized version of the k-anonymization problem.