Successful software evolves, more and more commonly, froma single system to a set of system variants tailored to meet the similiarand yet di erent functionality required by the distinct clients andusers. Software Product Line Engineering (SPLE) is a software developmentparadigm that has proven e ective for coping with this scenario. Atthe core of SPLE is variability modeling which employs Feature Models(FMs) as the de facto standard to represent the combinations of featuresthat distinguish the systems variants. Reverse engineering FMs consist inconstructing a feature model from a set of products descriptions. This researcharea is becoming increasingly active within the SPLE community,where the problem has been addressed with di erent perspectives andapproaches ranging from analysis of con guration scripts, use of propositionallogic or natural language techniques, to ad hoc algorithms. Inthis paper, we explore the feasibility of using Evolutionary Algorithms(EAs) to synthesize FMs from the feature sets that describe the systemvariants. We analyzed 59 representative case studies of di erent characteristicsand complexity. Our exploratory study found that FMs thatdenote proper supersets of the desired feature sets can be obtained with asmall number of generations. However, reducing the di erences betweenthese two sets with an e ective and scalable tness function remains anopen question.We believe that this work is a rst step towards leveragingthe extensive wealth of Search-Based Software Engineering techniques toaddress this and other variability management challenges.
展开▼
