The Use of Column-Based DBMSS to Form Multidimensional Data Structures

Abstract

The work is devoted to the analysis of the method of designing information systems using the “Data vault” modeling technology. The described approach can be used when building a data warehouse within a multidimensional information system based on the classical approach of data warehouse design – a 3-level data representation architecture, which includes data preparation area, or operational data warehouse, data warehouse, and thematic data marts. This approach allows you to organize data storage within a data warehouse using a metadata repository based on the multidimensional approach. The metadata repository is responsible for the process of collecting data, storing data, and presenting data for analysis. The “Data vault” approach allows you to design the data warehouse of an information system based on a metamodel that is semantically related to the domain of the system and is easily rebuilt it in the event of changes in the business model of the domain. When describing a metamodel, the first-order predicate calculus language is used, which makes it possible to control the metamodel using a declarative programming style – the “Prolog” language. The key point in the structure of the information system is the way of transition from the “Data vault” model to a multidimensional model of data representation based on associative rules of dependence between information objects.

Keywords

• data warehouse
• multidimensional data model
• data mart
• olap
• column-based dbms