Research On Multi-keyword Ranked Search Technology In Cloud Storage

In recent years,with the growing popularity of cloud storage technology,more and more users choose to outsource their data to public cloud servers.However,as a third-party platform,cloud storage service provider cannot be completely trusted.In order to guarantee the data privacy,users prefer to outsource their data in encrypted form in the cloud,which makes the traditional plaintext keyword search technique fail to encrypted data.Therefore,searchable encryption technology came into being which supports keyword retrieval over encrypted data,enabling efficient data utilization while protecting the privacy of user queries.The early searchable encryption schemes mostly support single keyword.In order to further improve the accuracy of retrieval,the multi-keyword ranked search over encrypted data technology has become one of the hot issues in the academic and industrial circles.Most existing multiple-keyword ranked search schemes over encrypted data do not satisfy required dynamic update,only a few are concerning about dynamic update which also have limitations with security and search efficiency.In addition,most existing retrieval schemes based on ciphertext only support exact or fuzzy match of keywords,ignoring the semantic relations among keywords,which may result in incomplete or inaccurate retrieval results.For the above deficiencies,the main results of the dissertation are summarized as following two aspects:1.This dissertation proposes a ranked search scheme over encrypted documents,which supports both multi-keyword and dynamic update.The solution uses a balanced binary tree as an index structure,supports parallel retrieval and dynamic update,and implements forward security and backward security during dynamic update.In order to achieve multi-keyword ranked retrieval,file index vector and query vector are constructed by vector space model and TF-IDF technique.To ensure the privacy of the data,the secure k-nearest neighbor algorithm is used to encrypt the file index and query vectors.2.On the basis of the first work,the semantic query technology is introduced.This dissertation proposes a multi-keyword ranked retrieval scheme over encrypted data that simultaneously supports semantic retrieval and dynamic update.In this scheme,the semanticextension of query keywords is implemented by constructing a semantic relationship graph.The performance analysis shows that the scheme has high retrieval efficiency.The security analysis shows that the scheme is security under the known ciphertext model.