A New Approach to Natural Language Query Search using Frequency Analysis Techniques in Cloud Computing

Abstract

Data outsourcing is the preferred act by owners of cloud data because of ease in maintenance. Data confidentiality of this outsourced sensitive data is a major task. Applying cryptographic techniques to outsourced data is the most secure way to achieve confidentiality. Searchable encryption techniques help in searching on encrypted data without actually decrypting it. These techniques limit the usefulness of data in the sense that searching encrypted data is difficult. With the increase in the volume of data, increases the size of indexing structure. This makes it more difficult to design cipher text based search scheme which facilitates reliable, memory efficient, fast retrieval on a huge volume of encrypted data. In this paper, keyword frequency based code method is presented to minimize the size of the index which makes fast retrieval of data inside a big data environment. The proposed system also supports secured, ranked retrieval using hash-based mapping structures. A hash-based technique efficiently retrieves the data using key-value pair data structure. The resulting system is able to handle queries which are written in natural language. Through extensive experiments using standard dataset, the performance of the system is validated. The system performance is evaluated and validated through extensive experimentation using standard dataset.  The results show that very less space for index storage is utilized by the system proposed in the paper which enhances the performance on the ground of the retrieval time.