Study On Multi-keyword Search Over Encrypted Cloud Data

Nowadays,as cloud computing develops,cloud storage service has become a popular approach to manage data for organizations and individuals,which is a component of cloud computing.However,cloud storage service is untrusted,which may leak data privacy.For some data owner,they are not willing to store their data,especially for some sensitive data such as financial records and personal emails into cloud servers on account of the data privacy.So it is necessary for data owners to encrypt their outsourced data.Unfortunately,the efficiency of the encrypted data utilization will be reduced and how to achieve keyword search over it becomes a problem.For the above issues,searchable encryption can provide some useful solutions.And it is of paramount importance to study multi-keyword search over encrypted cloud data,which is an important part of searchable encryption.In this paper,we conclude some related previous works and propose a secure multi-keyword search scheme over encrypted cloud data,which can support dynamic operation properly and effectively.To improve the search efficiency,we design a search index tree by utilizing Bloom filter.The algorithms of index tree construction and index tree query are presented.To achieve multi-keyword ranked search,our scheme utilizes the vector space model combined with TF×IDF rule and cosine similarity measure.Our scheme can achieve the sub-linear search time compared with other related work.What's more,the efficiency of dynamic operations in our scheme is better than other related schemes.The security analysis of our scheme shows that the basic scheme and enhanced scheme can meet the privacy requirements in the known ciphertext model and the known background model,respectively.Besides,the experiments indicate that our scheme is effective and feasible.