Analyzing the sensitivity of multi-objective software architecture refactoring to configuration characteristics

• DOI: 10.1016/j.infsof.2021.106568
• Publication Date: 2021-03-14T03:36:59Z
• Authors: Vittorio Cortellessa; Daniele Di Pompeo
• Keywords: Code refactoring

Software architecture refactoring can be induced by multiple reasons, such as satisfying new functional requirements or improving non-functional properties. Multi-objective optimization approaches have been widely used in the last few years to introduce automation process, and they revealed their potential especially when quantifiable attributes are targeted. However, effectiveness of heavily affected configuration characteristics algorithm, composition solutions. In this paper, we analyze behavior EASIER, which is an Evolutionary Approach for archItecturE Refactoring, while varying its characteristics, with objective studying find near-optimal solutions under different configurations. particular, use two solution space inspection algorithms (i.e., NSGA−II SPEA2) genome length composition. We conducted our experiments on a specific case study modeled Æmilia ADL, shown ability EASIER identify performance-critical elements software where worth applied. Beside this, from comparison multi-objective algorithms, has outperform SPEA2 most cases, although latter one able induce more diversity proposed Our results show that thoroughly automated process allows contexts evolutionary algorithm effectively finds Pareto