Design and Ex ante evaluation of an architecture for self-adaptive model-based DSS

Abstract

The quality of the decision support of a model-based Decision Support System (DSS) is fundamentally dependent on valid and actual models. A changing business environment can affect the validity of model components which could cause an incorrect model output. This problem is addressed in this paper by focusing on the self-adaptive property as a potential approach. To provide a model for decision support as close as possible to a dynamic business environment, the principles of self-adaptive systems are considered in an interconnected Model-/System-Controller (MoSyCo) architecture which is designed around DSS models. The design of the artifact is driven by a deduction of the problem characteristics to specify components of the intended architecture. The ex ante design evaluation is conducted in accordance to the stepwise evaluation by Sonnenberg and vom Brocke and considers a survey of 50 practitioners from the DSS domain.