An approach for source code classification using software metrics and fuzzy logic to improve code quality with refactoring techniques

Abstract

The problem of developing software quality is developer's experience which has different style and does not care about coding with principle that causes error or even bad smell. To reduce the risk of causing bad smell, the developer should concern with a good design principle and coding well. In addition, knowing the qualification and the characteristic of code is also important to promptly support verifying, recovering bad smell and improving them to be a good code. This research presents an approach for source code classification using software metrics and fuzzy logic to improve code quality with refactoring techniques. Our approach composed of 3 main sections; Source code definition with metrics and evaluation to classify source code type, Source code classification with fuzzy logic and Source code improvement with refactoring. The result of our approach is able to classify source code in type correctly and improve bad smell, ambiguous code to be a clean code. © 2011 Springer-Verlag.