Code quality improvement using Aquila Optimizer

Abstract

Abstract The software industry always requires high‐quality software. Code quality has been improved via refactoring and re‐refactoring and removing code smells. Refactoring is a technique to improve existing code by modifying inner structure, while code smells are indication that might have a wrong in code that causes a negative impact on the code quality. According to previous studies, the accuracy of code smell detection was not stable and could be improved by advanced algorithms. Additionally, the impact of refactoring on each of the quality attributes is unclear. This research proposes the use of Aquila Optimizer to detect code smells in addition to investigate the refactoring and re‐refactoring impact through ISO 9126 model. The experiment is conducted on five open‐sources Java projects to detect nine code smell types and measure the values of the ISO 9126 before and after the refactoring and re‐refactoring techniques to show the impact on code quality. The results showed that the average of detection accuracy is 95.18% and 95.68% for precision and recall, respectively. Moreover, the results in large projects are better than in the medium projects. The findings also show improvement in the values of the ISO 9126 model after applying refactoring and re‐refactoring.