Decision Support for the Selection of Commercial Off-The-Shelf (COTS) under Uncertainty

Developing software applications using Commercial Off-The-Shelf (COTS) software products has received much attention in the last decade. Several COTS selection methods have been proposed to select COTS products that best fit the requirements of software application. However, most current COTS selection methods only partially or not at all address information uncertainty related to COTS software products.;To validate UnHOS, we apply it to two case studies to select a COTS library software system and a database management system to develop an airline reservation system. The results show the potential advantage of using UnHOS over traditional COTS selection. The first case study shows that uncertainty representation influences the final COTS selection decision. Although uncertainty representation may not influence the COTS selection decision as in the second case study, uncertainty representation is still beneficial because it can be used to determine which criteria evaluators are not confident about the corresponding scores. These criteria are considered as sources of possible mismatches between system requirements and features of COTS candidates and they need significant attention from evaluators in particular if they are important to stakeholders.;In order to demonstrate the feasibility of tool support for UnHOS, the thesis proposes UnHOS-Tool. The UnHOS-Tool was used to apply UnHOS in the two case studies.;This thesis presents UnHOS, Uncertainty Handling aware COTS Selection, a COTS selection method distinguished from other COTS selection methods by its capability to represent uncertainty during COTS selection process. The UnHOS uses Analytic Hierarchy Process (AHP) to evaluate COTS candidates and Bayesian Belief Network (BBN) to represent uncertainty. In UnHOS, AHP is used to calculate overall score of COTS candidates. These scores reflect how well each candidate satisfies the evaluation criteria. BBN is used to represent uncertainty in terms of how much evaluators are confident about scores created by AHP. Scores and confidence values are used to make the COTS selection decision. This means, a COTS candidate is selected based on not only its score that represents its fitness with respect to the evaluation criteria but also the confidence level of the evaluation team.