A hierarchical sharing mechanism based privacy data protection
Zhaxi Namjie, Tianbo Lu, Yanjiao Gong, Jiaze Shang · 2021 3rd International Academic Exchange Conference on Science and Technology Innovation (IAECST) · 2021
The frequent occurrence of privacy data breaches and the subsequent issuance of data security laws and regulations have led to increased attention paid to the security of personal data. Ensuring the safety of personal privacy data while releasing data has become a significant challenge. Most of the existing privacy protection models provide the same query results for all data queries without considering the demand for data accuracy by different users, the privacy differences of data with different attributes, and the risk of privacy leakage caused by multiple queries. Based on this, this paper proposes a centralized differential privacy-based data hierarchical query algorithm from the data querier's demand for data precision in the data publishing scenario. Firstly, the corresponding trust level is calculated based on the three attribute values of the trust value of the data querier, the query authority level and the privacy level of the query data, and the privacy protection parameters corresponding to the trust level are used to realize the grading of the data querier. Secondly, the privacy budget parameters are set to avoid the risk of privacy leakage caused by multiple queries made by attackers. At the same time, the McSherry-Laplace mean algorithm is improved to reduce the error rate while ensuring that it satisfies differential privacy. The experimental results show that the above algorithm processes different datasets to provide accurate and differentiated usable data for various data querier users while satisfying differential privacy.