Privacy preserving attribute reduction for horizontally partitioned data

There has been concern over the apparent conflict between privacy and data mining. Attribute reduction is one of the most important contributions of rough set theory to data mining. In this paper, we address the issue of privacy preserving attribute reduction. Specifically, we consider a scenario in which two parties owning private data, wish to run a attribute reduction algorithm on the union of their databases, without revealing information about individuals. Our work is motivated by the need both to protect private information and to enable its use for research or other purposes. The above problem is a specific example of secure multi-party computation. We focus on the problem with the attribute reduction algorithm based on relative granularity, address an efficient protocol for securely computing the relative granularity, and present a privacy preserving attribute reduction algorithm for horizontally partitioned data.