Data Modeling in Big Data Systems Including Polystore and Heterogeneous Information Processing Components

This paper is studies modeling data in big data systems, including polystores and other heterogeneous information processing components. Currently, several works propose to harmonize polystore data models in this domain. This study considers various proposed methods; however, these solutions are not suitable for direct use for solving information security problems. Requirements on modeling the considered objects for solving security tasks and the level-sensitive modeling method based on the general security concept of polystores within a consistent approach are formulated. This study presents an authentic classification of the structure of data models of modern polystores and DBMSs, taking into account the mathematical framework in use. A new methodology of three-level modeling of data and processes in an object for protection is proposed; and the basics of models for all data representation levels are formulated. The results of this study lay the foundation for the integrated representation of data and processes for solving security problems and analyzing the security of big data systems.