Automating the choice of decision support system architecture

Abstract

Due to the wide-spread use of decision support systems (DSS), methods are required by software companies. Several concepts and methods have been suggested for decision-making. The development of DSS is still complex due to the variety of architectures, the number and the type of modules which compose them. Therefore, This paper present our method of DSS development, which is based on an architecture of 1 to 4 levels, focusing on the choice of an adapted architecture (composed of data warehouses and data marts). Our method is based on a mixed approach integrating the assessment of existing sources and user requirements. These are taken into account from the early step of DSS engineering, e.g. from the analysis step, so DSS architecture choice integrates them. © Springer-Verlag Berlin Heidelberg 2006.