Adapting Multidimensional Schemes to Data sources using Algebraic Operators (original) (raw)

2005, Software Engineering and Knowledge Engineering

Designing a decisional system requires a methodology different from those commonly adopted for operational information systems. In our methodology data marts are constructed on the basis of user requirements specified using OLAP design patterns. Since these patterns are independent of any data source, the data mart design process should solve the problems due to differences between user OLAP requirements, from one hand, and data source from the other hand. This paper, first defines multidimensional models. Secondly, it defines a set of operators for the validation and the refinement of patterns, covering the addition and deletion of: hierarchy, dimension, dimension attribute, Non dimension attribute, measure and fact.