会议专题

Multi-Objective Optimization of K in K-anonymity Model

  K-anonymity privacy model is a typical model to protect privacy when disseminating data involving individual subjects.The drawback of k-anonymity is that generalization will result in considerable loss of information.Further,under limited information loss an exhaustive analysis is required to determine a k value based on the privacy requirement by data publisher.Studies in this context have so far focused on minimizing the information loss for some given value of k,how to optimize k value that fits the specified data quality and privacy requirement is sparse.In this paper,we formulate a multi-objective optimization problem of k to illustrate that the decision of k can be much more exible.At first,one k bound is analyzed basing on the privacy disclosure,and another k bound is analyzed basing on the data quality metric.Then,the optimal k-value is gained by intersecting the two bounds.At last,an algorithm is employed to provide multi-objective optimization of k that is win-win on privacy requirement and data quality.

k-anonymity multi-objective data quality privacy requirement optimization

Jinling Song Liming Huang Chao Zhang Guangbin Zhang

Hebei Normal University of Science and Technology,Qinhuangdao 066004,China Environment Management College of China,Qinhuangdao 066004,China

国内会议

2014全国理论计算机科学学术年会

济南

英文

1-12

2014-10-16(万方平台首次上网日期,不代表论文的发表时间)