The maintainability of software depends on the quality of software. Software under evolution is modified and enhanced to cater the new requirements. Due to this the software becomes more complex and deviates from its original design, in turn lowering the quality. Refactoring makes object oriented software systems maintainable. Effective refactoring requires proper metrics to quantitatively ascertain the improvement in the quality after refactoring. In this direction, we have made an effort to quantitatively evaluate the quality enhancement by refactoring using dependency oriented complexity metrics. In this paper three experimental cases are given. The metrics have successfully indicated quantitatively the presence of defects and the improvement in the quality of designs after refactoring. These metrics have acted as quality indicators with respect to designs considered under ripple effects before and after refactoring. Metrics as quality indicators help to estimate the required maintenance efforts.
Published in:
Emerging Trends in Engineering and Technology (ICETET), 2009 2nd International Conference on
Date of Conference: 16-18 Dec. 2009