Thesis (M.Sc.) - Cairo University - Institute of Statistical Studies and Research - Department of Information Systems and Technology

Requirements Prioritization (RP) is considered one of the critical issues in requirements engineering (RE) process. It is usually treated as a Multi-Criteria Decision Making problem (MCDM). The main essence of RP is to help in implementing the core requirements of stakeholders with respect to cost, quality, available resources and delivery time. Due to information imperfections that normally exist in real world problems, RP is commonly handled as uncertain MCDM problem.Interval uncertainty is a type of uncertainties that represents uncertain score values as interval numbers. Ranking of interval numbers received much interest of researchers for many decades. In this thesis, a feasible and reliable ranking approach for interval numbers is proposed that can be adopted for RP problems. The proposed approach is based on the concept of penalizing uncertainty using a certain risk parameter.The proposed approach is dealing with two types of decision makers; Moderate Decision Maker (DM) and Conservative DM. The given illustrative examples ensure the efficacy and the competence of the proposed method compared with other methods existed in literature. Namely, it allows better handling of risk parameter according to the subjective view of the decision maker.The proposed approach is evaluated by solving examples existed in literature and comparing the results. It offers a simple and dynamic way to requirements engineers to prioritize software requirements under uncertainty

Issued also as CD