Research Of Multi-keyword Ranked Search Over Encrypted Cloud Data

With the rapid development of cloud computing,enterprises and individuals are more willing to store the encrypted documents to the cloud platform,thereby enhancing security and protect privacy effectively.However,the data encryption limits the ability of the user to perform keyword retrieval and makes the traditional plaintext retrieval method ineffective.Therefore,the effective ciphertext retrieval has become an urgent problem and the study of ciphertext retrieval technology is of great significance.This paper focuses on the multi-keyword sorting search of encrypted data in cloud platform and research on efficiency and security.The main work includes:In order to meet the needs of multi-keyword search for encrypted files,taking into account the privacy and data security of users under cloud server,we have improved a secure parallel index tree structure that supports multi-keyword search for encrypted data.Firstly,we use the index tree algorithm to construct the quadtree,the biggest advantage of index tree is that,with a smaller change in the structure of the tree to complete the update,on the basis of which to improve the dynamic update efficiency.We combine with the idea of public key encryption and user authentication information is added in the process of homomorphic encryption to solve the privacy leakage problem caused by the user's revocation.By classifying documents,similar documents can be stored centrally when searching for similarity ranking,so as to avoid unnecessary branch inquiries effectively and improve the query efficiency.In order to strengthen the privacy query,we propose a higher level of security model,when constructing of the keyword hash table,we add some random interference keywords to it.The improved method can well avoid the known plaintext attack.By adjusting the ratio of single document and total documents interference keywords,we can ensure the security while taking into account the efficiency and accuracy of the query.Finally,in order to further improve the accuracy of the query,we use the semantic-based Bayesian classification and it can be a good solution to the word polysemy and synonyms that can not match the exact documents.Finally,the complexity of the construction time of the index tree and the time complexity of the query are analyzed theoretically,meanwhile we discuss the different situations of the insertion and deletion of the document.Experiments onquery and update efficiency of index trees on 20news-18828 data set show that the query efficiency of the proposed method is greatly improved.Experiments on semantic-based Bayesian classification show that the proposed method can improve the query accuracy and reduce the false detection rate.