A benefit-driven genetic algorithm for balancing privacy and utility in database fragmentation
Yong-Feng Ge, Jinli Cao, Hua Wang, Jiao Yin, Wei–Jie Yu, Zhi‐Hui Zhan, Jun Zhang · Proceedings of the Genetic and Evolutionary Computation Conference · 2019
In outsourcing data storage, privacy and utility are significant concerns. Techniques such as data encryption can protect the privacy of sensitive information but affect the efficiency of data usage accordingly. By splitting attributes of sensitive associations, database fragmentation can protect data privacy. In the meantime, data utility can be improved through grouping data of high affinity. In this paper, a benefit-driven genetic algorithm is proposed to achieve a better balance between privacy and utility for database fragmentation. To integrate useful fragmentation information in different solutions, a matching strategy is designed. Two benefit-driven operators for mutation and improvement are proposed to construct valuable fragments and rearrange elements. The experimental results show that the proposed benefit-driven genetic algorithm is competitive when compared with existing approaches in database fragmentation.