“…Building a software product out of a particular SPL requires the selection of features that respect the desired software configuration. With the multiple characteristics/objectives that are interesting to decision-makers in practice (e.g., cost, technical feasibility, or reliability), the problem of finding the 'best' feature configuration is seen as an instance of a multi-objective optimisation problem [3,4].…”
Section: Introductionmentioning
confidence: 99%
“…,(3,4,3), (0.1,0.3,0.1), (1,2,1) and(3,4,3) for Correctness, Richness of features, Feature used before, Known defects, and Cost.-PROMETHEE-II: requires a parameter pair (weight and preference function)…”
Software Product Lines Engineering has created various tools that assist with the standardisation in the design and implementation of clusters of equivalent software systems with an explicit representation of variability choices in the form of Feature Models, making the selection of the most ideal software product a Feature Selection problem. With the increase in the number of properties, the problem needs to be defined as a multi-objective optimisation where objectives are considered independently one from another with the goal of finding and providing decision-makers a large and diverse set of non-dominated solutions/products. Following the optimisation, decision-makers define their own (often complex) preferences on how does the ideal software product look like. Then, they select the unique solution that matches their preferences the most and discard the rest of the solutions-sometimes with the help of some Multi-Criteria Decision Analysis technique. In this work, we study the usability and the performance of incorporating preferences of decision-makers by carrying-out Multi-Criteria Decision Analysis directly within the multi-objective optimisation to increase the chances of finding more solutions that match preferences of the decision-makers the most and avoid wasting execution time searching for non-dominated solutions that are poor with respect to decision-makers' preferences.
“…Building a software product out of a particular SPL requires the selection of features that respect the desired software configuration. With the multiple characteristics/objectives that are interesting to decision-makers in practice (e.g., cost, technical feasibility, or reliability), the problem of finding the 'best' feature configuration is seen as an instance of a multi-objective optimisation problem [3,4].…”
Section: Introductionmentioning
confidence: 99%
“…,(3,4,3), (0.1,0.3,0.1), (1,2,1) and(3,4,3) for Correctness, Richness of features, Feature used before, Known defects, and Cost.-PROMETHEE-II: requires a parameter pair (weight and preference function)…”
Software Product Lines Engineering has created various tools that assist with the standardisation in the design and implementation of clusters of equivalent software systems with an explicit representation of variability choices in the form of Feature Models, making the selection of the most ideal software product a Feature Selection problem. With the increase in the number of properties, the problem needs to be defined as a multi-objective optimisation where objectives are considered independently one from another with the goal of finding and providing decision-makers a large and diverse set of non-dominated solutions/products. Following the optimisation, decision-makers define their own (often complex) preferences on how does the ideal software product look like. Then, they select the unique solution that matches their preferences the most and discard the rest of the solutions-sometimes with the help of some Multi-Criteria Decision Analysis technique. In this work, we study the usability and the performance of incorporating preferences of decision-makers by carrying-out Multi-Criteria Decision Analysis directly within the multi-objective optimisation to increase the chances of finding more solutions that match preferences of the decision-makers the most and avoid wasting execution time searching for non-dominated solutions that are poor with respect to decision-makers' preferences.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.