Research On Encrypted Data With Keyword Ranked Search

With the rapid development of information technology,cloud services are gradually popularized due to their convenience and high cost-efficiency.More and more enterprises and individual users choose to store their data in the cloud,which has obviously become a trend.However,due to the consideration of privacy protection,users often need to encrypt their data before uploading it to the cloud to prevent the leakage of private information.Thus,a problem arises: how to combine data security with data retrieval? But the function design of the existing searchable encryption scheme is not comprehensive enough,often considering one thing and losing the other.Therefore,this paper aims to design an efficient and comprehensive encryption scheme using multi-keyword ranked search on the premise of ensuring data security.This thesis investigates the existing searchable encryption scheme.We have found that: the improvement of the search efficiency is the primary target of the existing multi-keyword ranked search technology;and the improvement of search function is not perfect which leaves a large space for development.The schemes of this thesis are mainly aims at such improvement: we take the keywords' semantic relationship between different dimensions into consideration: there exists approximate relationship between different dimensions of document index vector which represents different keywords,but the existing scheme treat this relationship as mutually independent.We improve the existing fuzzy-keyword matching scheme,and incorporateit into our scheme,so as to design a hierarchical semantic approximate multi-keyword ranked search scheme over Encrypted Data,which called HSAMS.The scheme supports exact matching,derivative-word matching and synonym matching of English keywords,corresponding to three-level index structure respectively,and supports flexible selection of three types of matching functions.The scheme also supports fuzzy search of misspelled query words.The security of the proposed scheme is analyzed,and the feasibility and efficiency of the proposed method is verified through comparative experiments.Because the document index in HSAMS scheme adopts the structure of hierarchical structure,the encryption stage will take a long time,so there is still a large space for improvement.The index structure and trapdoor matching mechanism are adjusted: the hierarchical index structure is eliminated,and a simplified scheme Semantic Approximate multi-keyword Ranked Search over Encrypted Data is designed,which called SAMS.When Data Owner(DO)expects to provide exact matching,derivative-word matching and synonym matching all three basic matching functions,SAMS is more efficient than HSAMS in index encryption stage,but it sacrifices flexibility in function selection.The safety,feasibility and high efficiency of SAMS were verified by experimental comparison between SAMS and HSAMS.