🤖 AI Summary
This paper addresses the Next Release Problem (NRP)—a multi-objective software requirements selection problem under resource constraints. We propose a scalable, generic optimization framework that uniformly models customer satisfaction, development cost, requirement attributes (e.g., priority, stability), inter-dependencies, and hard/soft constraints, enabling Pareto-optimal solution generation and stakeholder trade-off analysis. Our key contribution is the first formal, open-ended NRP modeling paradigm, designed to adaptively evolve with changing problem domains. Leveraging requirement dependency graphs, multi-objective optimization, and case-driven instantiation, we replicate and extend six existing solution approaches across six industrial case studies. Empirical results demonstrate the framework’s compatibility with diverse methodologies, high customizability, and practical effectiveness in real-world settings.
📝 Abstract
Due to the limited amount of resources available for the next release of the current product under development not all stakeholders requests can be included in the next product to deliver. This optimization problem, known as the Next Release Problem (NRP), has customers satisfaction and development costs as the basic optimization objectives, and has been the subject of many research works. However, there are additional issues that deserve to be considered and included in the definition of the NRP, such as supplementary optimization objectives including the elicited properties about the requirements, or the analysis of the non-dominated solution sets found to decide which solution is preferred. This paper presents a generic formulation for this problem that allows the management of the currently agreed properties and relationships between requirements. It provides an open formulation that is capable of growing as the scope of the NRP grows. We specify how our proposal can be used in software product projects when requirements selection has to be performed. We also describe how our formulation has been instantiated to cover previous solving approaches to this problem, using six case studies to demonstrate the successful customization of the generic formulation.