A Novel Method for protected ranked keyword search method in cloud

Abstract

Cloud data users prefer to outsource documents in an encrypted form for the purpose of preserving privacy. Therefore it is extremely important to develop efficient and reliable cipher-text search techniques. A major challenge is that the relationship between documents will be concealed in the process of encryption, which will lead to a significant degradation in search accuracy performance. In this paper, a hierarchical clustering method is proposed to support more search semantics and to meet the need for fast cipher-text search within a big data environment. The proposed hierarchical approach forms clusters of the documents based on a minimum relevance threshold, and later partitions the resulting clusters into sub-clusters until the constraint on the max size of cluster is reached. In the search phase, this approach does fairly good by reaching a linear computational complexity against an exponential size growth in the number of documents. In order to verify the authenticity and correctness of search results, minimum hash sub-tree is designed.