Multi Keyword Ranked Search for Secured Cloud Data Using Improved Vector Space Model

Abstract

The popularity of cloud computing has made data owners to outsource their data to cloud servers for great convenience and
reduced cost in data management. However, sensitive data should be encrypted before outsourcing for privacy
requirements, which obsoletes data utilization. We present a secured cloud data for rank based multi-keyword search with
chunking and replication. The vector space model and the QW-CDC models are combined in the index construction and
query generation. Each document is denoted by a vector, whose elements are the normalized QW values of keywords in
this document. Each query is also denoted as a vector, whose elements are the normalized CDC values of query keywords
in the document collection. The dot product of the QW vector and the CDC vector can be calculated to quantify the
relevance between the query and corresponding document. Ranking is done based on the relevance score obtained. We
construct a special KBB [Keyword Balanced Binary] tree structure and propose a “Greedy Depth-First Search” algorithm
to provide efficient multi-keyword ranked search. Thus the Multi-keyword ranked search provides the user with the most
relevant document and due to the tree based index structure the proposed scheme achieves sub-linear search time. The
secure DES algorithm is utilized to encrypt the documents for data security. Data is encrypted, split and stored in multiple
servers [chunking] which prevent the document from fully being hacked. Replication is done for data recovery and backup
which maintains the safety of the document.