Interactive Decision Support for Multiobjective COTS Selection

TL;DR: A decision support approach that involves first determining (feasible) Pareto-efficient alternatives, then allowing decision makers to interactively explore the solution space until they find the most appealing solution.

Abstract: In the past decade component-based software engineering (CBSE) has gained considerable attention from both industry and the scientific community. Obviously, selecting the "best" combination of commercial off-the-shelf (COTS) components plays a critical role in CBSE. This task becomes demanding, as multiple objectives and constraints have to be taken into account. So far, the selection process has been tackled by such traditional techniques as the weighted scoring method or the analytical hierarchy process. This paper introduces a decision support approach that more properly addresses that challenge. The approach involves first determining (feasible) Pareto-efficient alternatives, then allowing decision makers to interactively explore the solution space until they find the most appealing solution. It not only works without extensive a priori preference information (such as criteria weights), it can also be easily integrated into existing COTS selection frameworks