A Study on Privacy Preservation for Multi-user and Multi-granularity

Dong Li,Xianmang He,Huahui Chen,Yihong Dong,Yefang Chen
DOI: https://doi.org/10.1109/icdmw.2013.31
2013-01-01
Abstract:Currently, all the existing studies with good privacy guarantees focus on a single privacy level. Namely, a certain degree of privacy protection is implemented on all anonymized data released. However, this is not consistent with the actual scene that the different roles have different levels of privacy. From this point of view, this paper proposed a scenario with multi-user and multi-granularity privacy protection, and proposed the l-increment privacy protection model. On this basis, we put forward a generalization algorithm, which can meet the requirement for multi-user and multi-granularity, and reduce greatly the amount of information loss resulting from data generalization for implementing data anonymization in the meanwhile. Our findings are verified by experiments.
What problem does this paper attempt to address?