A hierarchical clustering method for big data oriented ciphertext search

Abstract

Following the wide use of cloud services, the volume of data stored in the data center has experienced a dramatically growth which makes real-time information retrieval much more difficult than before. Furthermore, text information is usually encrypted before being outsourced to data centers in order to protect users' data privacy. Current techniques to search on encrypted data do not perform well within such a massive data environment. In this paper, a hierarchical clustering method for ciphertext search within a big data environment is proposed. The proposed approach clusters the documents based on the minimum similarity threshold, and then partitions the resultant clusters into sub-clusters until the constraint on the maximum size of cluster is reached. In the search phase, this approach can reach a linear computational complexity against exponential size of document collection. In addition, retrieved documents have a better relationship with each other than traditional methods. An experiment has been conducted using the collection set built from the recent ten years' IEEE INFOCOM publications, including about 3000 documents with nearly 5300 keywords. The results have validated our proposed approach.