Reducing maintenance efforts of developers by prioritizing different code smells

Abstract

An architectural problem associated with a software system constantly affects the evolving system. These architectural problems are symptoms of different code smells that must be removed using refactoring. The refactoring efforts are directly associated with the maintenance cost of the software system. This cost can be minimized or optimized by prioritizing different code smells. Prioritization helps tackle only a subset of code smells and hence minimize the maintenance cost. This makes code smells ranking and prioritization an important research area and is tackled in this paper. This paper proposes a new approach that is capable of ranking the existing code smells. This prioritization is based on the newly proposed metric based on three identified key criteria. Firstly, the severity of a code smells based on the change-history of a software system. Secondly, the association of the code smells with the improvement in the understandability of the software system. Thirdly, the importance of a developer’s feedback for a given code smells associated with a class in the software system. The feasibility of the proposed approach is tested and evaluated on an open-source Java software system.