Quality Metrics for Data Warehouse Multidimensional Models with Focus on Dimension Hierarchy Sharing

Abstract

Data warehouses, based on multidimensional models, have emerged as powerful tool for strategic decision making in the organizations. So it is crucial to assure their information quality, which largely depends on the multidimensional model quality. Few researchers have proposed some useful metrics to assess the quality of the multidimensional models. However, there are certain characteristics of dimension hierarchies (such as relationship between dimension levels; sharing of some hierarchy levels within a dimension, among various dimensions etc.) that have not been considered so far and may contribute significantly to structural complexity of multidimensional data models. The objective of this work is to propose metrics to compute the structural complexity of multidimensional models. The focus is on the sharing of levels among dimension hierarchies, as it may elevate the structural complexity of multidimensional models, thereby affecting understandability and in turn maintainability of these models.