Ciphertext Ranked Search Scheme Based on Dictionary Classification and Attribute Weighting
Searchable encryption supports users to perform search operations over encrypted data without decrypting the original data.The time cost of index and trapdoor construction of existing multi-keyword ranked searchable encryption schemes usually depends on the vector space formed by the global keyword dictionary.To reduce computation overhead and communication cost on the users side and further enhance users'satisfaction with search results,this paper proposes a multi-keyword ranked search scheme that supports fine-grained access control.The scheme first designs a dictionary-stripping mechanism based on mutual information,through which the keywords in the global dictionary are differentiated into two subsidiary sub-dictionaries with different information entropy,which further generates indexes and trapdoors in the low-dimensional sub-dictionary space.Secondly,the weight of attributes in the document access policy is considered as one of the ranking criteria so that data users achieve more relevant results.Finally,the filtering vector filters the data for the first round,and the attribute matching is used to complete the second round of elimination to avoid unnecessary computation during the search.
万方数据
维普
