Data modeling for consistent access control in heterogeneous big data systems

The evolution of data management systems, both in terms of the volume of information and the number and diversity of components, has reached a new stage. Technologically, it is characterized by the combination of heterogeneous tools in a single big data system with a common information lifecycle. The use of a unified information processing ecosystem requires a unified approach to solving security problems, including access control. Today, this is a difficult task because data tools use not only different security models, but also different data structures. At the same time, the creation of a unified approach is impossible without a unified representation of data and operations across the processing system. The paper considers the problem of coordinated access control in big data systems, requirements for its implementation. The analysis of existing methods and models of integration of heterogeneous data processing tools shows the need to continue work on a universal data model. The authors propose a universal model based on aggregates and set theory. The new model is more universal than the previously proposed ones by expanding the list of data relations and operations. Mapping techniques between the new model and DBMS models in the big data ecosystem are also provided. The proposed solution is a step towards the creation of big data security technology within the framework of a consistent security approach..