Design,optimization And Application Of Searchable Encryption With Multi-keyword And Multi-function In Cloud Environment

In recent years,with the rapid development and popularization of cloud computing,more and more companies and individuals have begun to outsource data or computation to the cloud server.This kind of service model can not only be used on demand or pay-per-use,but it is also more convenient to expand and maintain.However,outsourced data such as e-mail or medical records may leak users' privacy.In order to protect the confidentiality of the users' data,the traditional method is to encrypt the data before outsourcing it to the cloud server.But the encryption will greatly limit the availability of data owing to the processing method applicable to plaintext can't always be applied directly to ciphertext.Symmetric searchable encryption,a promising method,guarantees the confidentiality of the data and supports search directly on encrypted data.Currently,the research hotspots of symmetric searchable encryption have gradually shifted from single-keyword searchable encryption to multi-keyword searchable encryption or even functional searchable encryption.However,it still needs further research on how to support functional searchable encryption while improving the efficiency,and how to apply functional searchable encryption to large-scale document collections.To solve the problems above,we firstly propose a basic scheme for top-k ranked and verifiable multi-keyword searchable encryption that supports preference search and boolean search.The basic scheme constructs a corresponding trapdoor to support preference search and boolean search respectively.It also improves the efficiency of top-k ranking search by establishing a tree-based index and employing the pruning strategy.Furthermore,the validation object is constructed to ensure that the search results are not tampered or disordered during transmission.On the other hand,large-scale document collections tend to generate a highdimensions keyword dictionary,which can significantly affect the efficiency in all phases of searchable encryption.To the end,we propose three optimized schemes,including a scheme based on keyword extraction and automatic text summarization,a scheme based on stemming and lemmatization,and a scheme based on the diagonal matrix.Those schemes make it possible to support large-scale document collections.We briefly analyze the security of the proposed schemes and conduct a series of experiments to quantificationally compare the efficiency between our schemes and the other schemes at each stage.The experimental results demonstrate the feasibility and efficiency of the proposed basic scheme.In addition,we also conduct a series of experiments of the preference search under the basic scheme and three optimized schemes on a real-world data set.The experimental results show that three optimized schemes have a considerable degree of efficiency improvement compared with the basic schemes at each stage.Finally,we discuss several possible application scenarios of searchable encryption to illustrate application prospects of searchable encryption.