Private and Secure Secret Shared MapReduce (Extended Abstract)

摘要: Data outsourcing allows data owners to keep their in public clouds, which do not ensure the privacy of and computations. One fundamental useful framework for processing a distributed fashion is MapReduce. In this paper, we investigate present techniques executing MapReduce computations cloud while preserving privacy. Specifically, propose technique outsource database using Shamir secret-sharing scheme then, provide privacy-preserving algorithms performing search fetch, equijoin, range queries Consequently, our proposed algorithms, cannot learn or All eliminate role owner, only creates distributes secret-shares once, minimize user, needs perform simple operation result reconstructing. We evaluate efficiency by (i) number communication rounds (between user cloud), (ii) total amount bit flow (iii) computational load at user-side cloud-side.