No more privacy Concern: A privacy-chain based homomorphic encryption scheme and statistical method for privacy preservation of user's private and sensitive data

Internet giants collects a huge volume of data using the modern technologies and apply them to data mining for prediction and decision making in numerous fields, for instance, health care, web analysis, insurance, market basket analysis and bioinformatics. However, the huge repositories collects variety of data, it inevitably gathers a mass of private and sensitive data about the users. The revelation of such private and sensitive information may harm the user's privacy. In recent years, privacy is becoming an increasingly main concern in many data mining applications. Privacy preserving data mining (PPDM) has been evolved as the solution for the privacy issues. PPDM deals with protecting the individual's data without compromising the data utility. In this paper, privacy-chain based homomorphic encryption scheme along with the statistical transformation method weight of evi-dence and information value is proposed to preserve the private and sensitive information about the user. The proposed statistical transformation with homomorphic encryption (STHE) algorithm will perturb both the nu-merical and categorical values from the adult income, bank marketing and lung cancer datasets without affecting the data utility. In STHE, the quasi-identifiers are initially perturbed with IV and then privacy-chain based ho-momorphic encryption using RSA is applied to preserve the data privacy. The performance of the STHE algorithm is made compared with the state-of-the-art algorithms using the classifier models, decision tree, random forest, extreme gradient boost, and support vector machines. The experimental results shows that the proposed STHE algorithm outperforms the existing techniques in terms of accuracy, performance, data transformation, data retrieval and privacy preservation level along with the data utility.

Automated summary

Internet giants collect a large amount of data and use them in data mining for prediction and decision making in various fields. However, this collection of data inevitably includes private and sensitive information about users, raising concerns about privacy. Privacy preserving data mining (PPDM) has emerged as a solution for this issue. This paper proposes a privacy-chain based homomorphic encryption scheme combined with statistical transformation methods to preserve private and sensitive information without compromising data utility.