Abstract：Data fusion release can provide users with more effective statistical results and more personalized services.The existing data fusion model mainly adopts kanonymity privacy protection method,which does not take into account the impact of data user's authority and importance on data privacy,resulting in the serious leakage of privacy data and a serious decline in data quality.To solve this problem,a data hierarchical fusion publishing mechanism based on differential privacy protection(HDFPM)is proposed.This mechanism classifies users' rights and payment.In the process of data fusion,differential privacy protection technology is combined with the classification tree and its improved algorithm to perform data fusion,and the hierarchical differential privacy budget is reasonably distributed to realize the classification of fusion data privacy protection.Experimental results show that this mechanism can not only realize effective data fusion,but also protect sensitive data.
李万杰,张兴,曹光辉,李帅,张青云. 基于差分隐私保护的数据分级融合发布机制[J]. 小型微型计算机系统, 2019, 40(10): 2252-2256.
LI Wan-jie,ZHANG Xing,CAO Guang-hui,LI Shuai,ZHANG Qing-yun. Hierarchical Data Fusion Publishing Mechanism Based on Differential Privacy Protection. Journal of Chinese Computer Systems, 2019, 40(10): 2252-2256.
2019, Vol. 40 
