Approaches to the development of multi-dimensional databases: lessons from four case studies

The paper explores the manner in which an organization's data and information can be effectively utilized to assist an organization to achieve its business objectives. With the increased popularity of data warehousing and executive information systems, there is renewed interest by IT practitioners in data models and database structures, in particular multi-dimensional forms, which have joined their relational counterparts as legitimate tools for extracting vital business information from an organization's operational data. As much of the current literature on multi-dimensional databases comes in the form of white papers from the creators of multi-dimensional systems, there is a need for a more objective and theoretical analysis of this area, particularly on aspects that concern application developers and end users. To this end, the paper presents an analysis of four case studies involving the development of multidimensional databases applications. The initial two cases used different approaches, one top-down and one bottom-up. The successful aspects of these cases were used to create a "middle-out" approach that guided two subsequent case studies, thereby demonstrating that the dimensional view of data does provide managers with an effective means of making sense of organizational data.