1. A novel ε-sensitive correlation indistinguishable scheme for publishing location data
- Author
-
Zhang Lei, Zhang Guo-yin, and Wang Bin
- Subjects
Computer and Information Sciences ,Computer science ,Science ,Entropy ,Social Sciences ,02 engineering and technology ,Research and Analysis Methods ,computer.software_genre ,Access to Information ,Sociology ,0202 electrical engineering, electronic engineering, information engineering ,Data Mining ,Humans ,Cell Cycle and Cell Division ,Computer Security ,Data Processing ,Multidisciplinary ,Applied Mathematics ,Simulation and Modeling ,Physics ,Quality of service ,Biology and Life Sciences ,Social Communication ,020206 networking & telecommunications ,Cell Biology ,Mobile Applications ,Communications ,Data Reduction ,Cell Processes ,Privacy ,Physical Sciences ,Remote Sensing Technology ,Location-based service ,Telecommunications ,Medicine ,Thermodynamics ,Engineering and Technology ,020201 artificial intelligence & image processing ,Data mining ,Information Technology ,computer ,Mathematics ,Algorithms ,Research Article - Abstract
Nowadays, location based service (LBS) is one of the most popular mobile apps and following with humongous of location data been produced. The publishing of location data can provide benefit for promoting the quality of service, optimizing the commercial environment as well as harmonizing the infrastructure construction. However, as location data may contain some sensitive or confidential information, the publishing may reveal privacy and bring hazards. So the published data had to be disposed to protect the privacy. In order to cope with this problem, a number of algorithms based on the strategy of k-anonymity were proposed, but this is not enough for the privacy protection, as the correlation between the sensitive region and the background knowledge can be used to infer the real location. Thus, consider about this condition, in this paper a ε-sensitive correlation privacy protection scheme is proposed, and provides correlation indistinguishable to the location data. In this scheme, entropy is first used to determine the location centroid of each cell to build up the voronoi diagram. Then the coordinate of the untreated location data that is located in the cell is transferred into the centroid vicinity. Accordingly, the sensitive correlation is destroyed by the coordinate of each published data. The process of transferring the location data is determined by metrics of ε-sensitive correlation privacy, and is rigorous in mathematical justification. At last, security analysis is proposed in this paper to verify the privacy ability of our proposed algorithm based on voronoi diagram and entropy, and then we utilize the comparative experiment to further affirm the advantage of this algorithm in the location data privacy protection as well as the availability of published data.
- Published
- 2019