An Information Loss Reduction Scheme in Big Data k-Anonymization

초록

Information loss is one of the critical issues to be resolved when applying k-anonymization to the publishing data. To solve the problem, several solutions has been proposed by many researchers. However, the existing approaches cannot be used almost for BigData because it takes too long time. For Big Data, the computational time required to reduce the information loss is regarded as a NP-hard problem. To deal with the limitation, this paper presents a scheme that is based on heuristic threshold definition. To evaluate the proposed approach, we have implemented a pro-type evaluation system. The experimental results shows that our approach improve the execution time a little for Big Data when compared with the existing approaches.

키워드