CRSA cryptosystem based secure data mining model for business intelligence applications

Abstract

In the present competitive scenario, business intelligence (BI) applications play a very significant role for organizational decision support system (DSS). These BI applications need huge data transactions and information sets along with certain efficient data mining model to present certain expected result visualization. Data warehouses and OLAPs are the key components for BI tools in which information are collected from varied data sources, alteration or even certain improper data might lead wrong decision making, resulting into devastating consequences. Therefore, security and authenticity of datasets are must in such applications. In this paper, decision tree algorithm C5.0 has been employed for data mining and to ensure data authenticity, a Commutative RSA (CRSA) based privacy preserving model has been developed. The proposed C5.0 algorithm based privacy preserved and secure mining model has exhibited not only the optimum security of data or user associated datasets but it has also optimized mining efficiency in terms of higher accuracy, minimum execution time and lower key computation and distribution overheads.