Secure Top-k Search Over Encrypted Data In Outsourcing Cloud Environment

With the maturity and popularity of cloud computing,an increasing number of enterprises and individuals tend to outsource their storage and computing resources to cloud servers.This behavior can not only save costs but also makes it easy to use resources anytime and anywhere.However,while enjoying the large-scale and efficient services provided by cloud servers,users also face the danger of leaking sensitive information of outsourced data,such as personal logs,financial transaction files and so on.To minimize the risk of leaking outsourced data,data owners want to encrypt sensitive data before outsourcing.Although encryption can improve the security of data,it also affects the availability of data.It is necessary to design corresponding security mechanisms to solve this problem.At present,as an essential way to realize security keyword search in the cloud environment,the searchable encryption has become a research hotspot for the protection of data security without affecting the practicability of data.In view of the security issues faced by keyword search in the cloud environment,this paper proposes three different schemes from the perspectives of privacy and consistency.(1)Privacy-preserving Top-k Search(PTS): Based on the inverted index,the scheme divides the posting list into equal lengths according to the partition parameter,and constructs corresponding block vectors for each block generated.Then,the block vectors are used as the new terms of linear partition inverted index,and the corresponding blocks are used as the posting lists.Finally,the complete binary tree is used to transform the linear structure into the tree structure and generate the tree partition inverted index.Based on the tree partition inverted index,a "secondary filtering" algorithm is proposed,which can quickly filter the blocks without search keywords and greatly improve the search efficiency.(2)Verifiable Privacy-preserving Top-k Search(VPTS): Considering that the cloud server may have malicious attacks on its own,the ciphertext set returned to the authorized user may be incomplete or incorrect.To this end,a security Top-k search scheme for integrity verification is proposed.The scheme is also based on the inverted index.It constructs partial order constraint chain according to the correlation between keywords and documents.Authorized users can reconstruct the partial order constraint chain to verify the correctness of the results after obtaining the verification information returned by the cloud server.(3)Parallel Privacy-preserving Top-k Search(PPTS): In order to embody the characteristics of cloud computing such as distributed,large-scale and dynamic,a parallel privacy-preserving Top-k search scheme is proposed by using the traditional Map-Reduce distributed computing framework.The scheme can run directly on the Hadoop distributed platform and process the search requests of authorized users in parallel.It can greatly improve the ability of cloud server to process data and provide search services and is more suitable for large-scale document collection.