Multidimensional Modelling in NoSQL Database : A Systematic Review

Abstract

NoSQL Database Manager Systems (DBMS) are an alternative to analytical systems and have been used in this regard. As analytical systems often use data multidimensional modelling, the question arises of how to apply this modelling to NoSQL DBMS, which has such differentiated physical characteristics of relational databases. A systematic literature review (SLR) is performed to gather the aspects prompting the application of multidimensional modelling to NoSQL. One hundred twelve were pre-selected, and 67 were classified as relevant according to the research questions defined for this SLR. Three main approaches are highlighted: classical multidimensional modelling, multidimensional modelling with junctions of aggregated facts and dimensions in a single entity, and multidimensional modelling with aggregated facts and dimensions in a single entity, and adding pre-computed summaries. These approaches are explained, and their differences and similarities are detailed.