Diverse multi-keyword ranked search over encrypted cloud data supporting range query

Abstract

With the development of cloud computing over encrypted outsourced data, textual multi-keyword search schemes over encrypted data have been proposed. In fact, with various types of data available, keyword searches can not only be texts but also any other digital information, such as location, time, and price, etc. Currently, diverse search, which combines textual and digital keywords, has been studied over plaintext, i.e., unencrypted cloud data. However, it is challenging to develop a diverse search over encrypted cloud data. To address this issue, in the paper, we leverage the k-nearest neighbor (kNN) technique to design a diverse multi-keyword ranked search scheme over encrypted cloud data. Specifically, we put the digital keywords and textual keywords together in one index and use only one set of keys to encrypt the combined index. Security analysis indicates that our proposed scheme can achieve textual keyword privacy, digital keyword privacy and trapdoor unlinkability. Functionality analysis shows that our proposed scheme can achieve textual keyword search, coordinate matching, digital keyword search and range query. Performance evaluation demonstrates that our proposed scheme is efficient in terms of index construction, trapdoor generation and query.