Cost-benefit analysis of data warehouse design methodologies

Methodologies for data warehouse design are increasing more and more in last years, and each of them proposes a different point of view. Among all the methodologies present in literature, the promising ones are the hybrid methodologies—because they represent the only way to ensure a multidimensional schema to be both consistent with data sources and adherent to user business goals—and those able to support the designer by providing some kind of automation. However, the results obtainable by the methodologies can differ substantially in terms of schema quality and required efforts. In this paper, we provide metrics for evaluating the quality of multidimensional schemata in reference to the effort spent in the design process and the automation degree of the methodology. As a case study, we apply our evaluation to the major emerging hybrid methodologies for data warehouse schema design.