Location Based Privacy Preserving Access Control for Relational Data

摘要

Rapid expansion of network and internet services enabled users to use and share large amount of data on a massive scale. Once the information is combined, it becomes the wealth information which can be used for research. Researcher directly applies data mining techniques and algorithm on the original dataset to fetch information, which may leads to leakage of privacy data. Large amount of data leads to exposure of identity. To meet this privacy concern, unique identity is removed from the original data before applying publishing data for research. Even though individual identity is disclosure by linking different datasets. To protect privacy, privacy preserving mechanism (PPM) is used. In this paper, we suggested new method to get desired level of privacy stored in both local and distributed environment. Our methodology for privacy includes anonymization technique applied to grouped data so as to get more accuracy. In this proposed method, we applied generalization privacy technique to selected quasi identifier by setting range values as min-max. Further published data contains only two records of each group with their respective counts, instead of publishing repetitive records, in order to increase the performance in the distributed environment. In addition to that for extra Security, access control is achieved by the location.