A secure Search over Distributed Data

Abstract

In recent years, due to the economic benefits and technical advances of cloud
computing, huge amounts of data have been outsourced in the cloud. To protect the
privacy of their sensitive data, data owners have to encrypt their data prior
outsourcing it to the untrusted cloud servers. To facilitate searching over encrypted
data, several approaches have been provided. However, the majority of these
approaches handle Boolean search but not ranked search; a widely accepted
technique in the current information retrieval (IR) systems to retrieve only the top–k
relevant files. In this paper, propose a distributed secure ranked search scheme over
the encrypted cloud servers. Such scheme allows for the authorized user to search
the distributed documents in a descending order with respect to their relevance
documents and the contents of files be retrieved. To do so, each data owner builds
its own searchable index, and associates with each document in that index its weight
score, which facilitate document ranking. To ensure the privacy of these weights,
utilized the ranking. While preserving their capability to perform the ranking process
and to conducted several empirical analyses on a real dataset, all programs needed to
the propose system written using Matlab language.