Efficient privacy-preservation multi-factor ranking with approximate search over encrypted big cloud data

Abstract

Encrypting data before outsourcing data has become a challenge in using traditional search algorithms. Many techniques have been proposed to cater the needs. However, as cloud service has a pay-as-you-go basis, these techniques are inefficiency. In this paper we attack the challenging problem by proposing an approximate multi keyword search with multi factor ranking over encrypted cloud data. Moreover, we establish strict privacy requirements and prove that the proposed scheme is secure in terms of privacy. To the best of our knowledge, we are the first who propose approximate matching technique on semantic search. Furthermore, to improve search efficiency, we consider multi-factor ranking technique to rank a query for documents. Through comprehensive experimental analysis combined with real world data, our proposed technique shows more efficiency and can retrieve more accurate results and meanwhile improve privacy by introducing randomness in query data.