Multi-Objective Optimization of Feature Model in Software Product Line: Perspectives and Challenges

Abstract

Software Product Line (SPL) is process for developing families of software with reusability of features categorized as common and variable features. Feature Model (FM) is developed to manage these features. Common features are easy to manage, however variable features are hard to manage because of complex relations and constraints between features. Optimization is required to manage the variabilities for best selection of features and product configurations. To this end, different Multi-Objective Evolutionary Algorithms have been proposed to get the optimal solutions of feature model. In this paper we have compared among three main optimization algorithms i.e. IBEA, NSGA-II and MOCell. Our comparison is based on previous research correctness solutions for product뭩 configuration with five objective functions on different feature models from SPLOT and LVAT repositories. The goal of this comparison is to find the current research prospective and challenges of multi-objective optimization in FM.