辽宁工程技术大学学报（自然科学版）2024，Vol.43Issue(4)：506-512,7.DOI:10.11956/j.issn.1008-0562.20230180
基于时空特征的差分隐私保护方法
Differential privacy protection method based on spatiotemporal correlation
摘要
Abstract
In order to better protect the personal sensitive information embedded in trajectory data,an approach builds upon existing privacy protection frameworks and further investigates the correlation between temporal characteristics and spatial features is proposed by combining theoretical research with empirical validation,which named differential privacy protection method based on spatiotemporal correlation features(SC-DP).The method adaptively allocates weights of time and space dimensions according to the local dynamic of trajectories,ensuring the sub-stream appearing in a continuous ε-time window and m×m regions that can meet the requirements of ε-differential privacy protection.The algorithm's feasibility and practicality were validated on real datasets,achieving a 32.8%reduction in average error in the Geolife dataset and a 29.2%performance improvement in the CabSpotting dataset.关键词
时空关联特征/轨迹子流/位置服务/差分隐私/数据可用性Key words
temporal and spatial correlation characteristics/sub-stream of trajectory/location services/differential privacy/data availability分类
信息技术与安全科学引用本文复制引用
乔雨,嵇浩..基于时空特征的差分隐私保护方法[J].辽宁工程技术大学学报（自然科学版）,2024,43(4):506-512,7.基金项目
江苏省高等学校基础科学(自然科学)研究项目(21KJB520028) （自然科学）
南京工业大学浦江学院自然科学立项项目(njpj2022-1-28) （njpj2022-1-28）
