SECURE KEYWORD SEARCH IN CLOUD STORAGE USING SECRET SHARING WITH MACHINE LEARNING-ENHANCED ACCESS CONTROL

Abstract

Cloud computing is an important data storage and sharing platform. Nonetheless, it is a tough challenge to preserve the privacy of data and at the same time, search through encrypted data efficiently. The paper provides a more secure key-word search model which combines (k, n) secret sharing with query optimization by using machine learning and lightweight encryption. The suggested system spreads encrypted file shares among a variety of cloud servers, where one server cannot recreate stored information. The keyword indexing is done using TF-IDF in conjunction with Logistic Regression to speed up document retrieval. The control of access is done by means of secret share verification that prevents unauthorized access. Experimental findings show that query time and encryption computation overhead are reduced by 38 and 45 percent respectively in comparison to the traditional searchable encryption systems that are based on AES. “The architecture balances well in terms of security, usability and real-time performance of large-scale cloud applications.