Facilitating Efficient Encrypted Document Storage and Retrieval in a Cloud Framework

Abstract

In the recent years, cloud computing has gained a lot of attention and big IT giants are providing cloudservices to the clients. As a part of this service user data are stored in remote server machines which are managed by the cloud service provider.This necessitates the encryption of user’s sensitive information before outsourcing to the commercial public cloud. Huge amount of information present in the cloud calls for an efficient data retrieval mechanism.We are proposing a novel approach for keyword based search on the encrypted cloud data which facilitates easy retrieval of information from multiple encrypted data sources. Some of the existing approaches for the keyword based search focuses on Boolean search which will consider only the presence of the keyword in the file collection. Some other methods has server-side ranking wherethe encrypted relevance score may maintain some specific order which may lead to leakage of information. This paper uses latest information retrieval procedures along with vector space model and homomorphic encryption to ensure the search accuracy and secrecy while concealing the access pattern and search pattern. It provides a multiuser environment that supports storage of data belonging to multiple data owners andsearch request from multiple users. A salient feature of the suggested method is it allows efficient update operation of the file collection by the data owners. This process avoids the overhead of uploading/calculating the index scores for each of the existing documents from scratch for any file collection change.