Medical data publishing based on average distribution and clustering

Abstract

Most of the data publishing methods have not considered sensitivity protection, and hence the adversary can disclose privacy by sensitivity attack. Faced with this problem, this paper presents a medical data publishing method based on sensitivity determination. To protect the sensitivity, the sensitivity of disease information is determined by semantics. To seek the trade‐off between information utility and privacy security, the new method focusses on the protection of sensitive values with high sensitivity and assigns the highly sensitive disease information to groups as evenly as possible. The experiments are conducted on two real‐world datasets, of which the records include various attributes of patients. To measure sensitivity protection, the authors define a metric, which can evaluate the degree of sensitivity disclosure. Besides, additional information loss and discernability metrics are used to measure the availability of released tables. The experimental results indicate that the new method can provide better privacy than the traditional one while the information utility is guaranteed. Besides value protection, the proposed method can provide sensitivity protection and available releasing for medical data.

A privacy‐preserving method for publishing data with multiple sensitive attributes

“…The multiple SAs data publishing methods have different concerns, even though they aim to balance the privacy protection and information utility. Refer to our previous work [9], the previous data publishing methods for multiple SAs [4,12,13,[17][18][19][20][21][22][23][24][25][26][27][28] can be divided into two categories according to their concerns: protection-oriented methods and utility-oriented methods.…”

“…It is clear that L-diversity [3] ensures the diversity of SA values in a group, and limits the frequency of each SA value. Most of the subsequent privacy models [5][6][7][8][9][10][11] are based on the frame of K-anonymity [1] or L-diversity [3]. However, these studies focus on a single SA and the practical application usually involves multiple SAs.…”

A multi‐feature‐based intelligent redundancy elimination scheme for cloud‐assisted health systems

“…Related work reveals that there is almost no redundancy in the data of different types [22,39]. Thus, we classify medical data according to the data type and then test the redundancy of multimodal medical data.…”

A multiple sensitive attributes data publishing method with guaranteed information utility

“…The goal of data publishing is to guarantee the information utility with privacy protection. Most of the previous methods [1][2][3][4][5][6][7][8] are based on K-anonymity [9] and L-diversity [10], where multiple QIs and a single SA are included. To handle the scenario of multiple sensitive attributes, a series of approaches for multiple SAs data publishing have been presented [11][12][13][14][15][16][17][18][19][20][21] in recent years.…”