The Generic Framework of Privacy Preserving Data Mining Phases: Challenges & Future Directions

Abstract

The recent era is witnessing great advances in technologies, communication, and storage methods. This gave the ability to business organizations to collect and store huge volumes of data about their business and individuals. In addition, data mining has gained more attention and usability in many business fields to extract useful information and insights, discovering unknown patterns from such massive, collected data. However, in many domains, these data include private information about people and cannot be utilized and shared directly due to several individuals’ privacy issues. Hence, Privacy Preserving Data Mining (PPDM) has received a huge consideration in the current research to utilize the private collected data in different data mining tasks, while protecting the individuals’ privacy. Various review studies for PPDM have been proposed in the literature, presenting different types of techniques used to preserve data privacy. However, no systematic description for the involved steps that are performed thoroughly in each PPDM technique was presented. In this paper, we present a general deduced framework for PPDM, clarifying the different phases performed through the associated processes. Besides, based on the phase at which the technique can be applied, a classification for the different types of PPDM techniques proposed in the literature is provided. Finally, concerns that need to be addressed and future research directions in the PPDM research field are highlighted.