Application of a Fuzzy Inference System to Measure Maintainability of Object-Oriented Software

In the software development life cycle, maintenance is the most costly activity because it requires more effort compared to other activities. To reduce the cost of software maintenance, it is essential to predict software maintainability during the early phases of software development. As a consequence of early estimation, further corrective and preventive actions can be performed more efficiently to improve the maintainability of the software. Predicting maintainability of software using fuzzy logic is gaining more attention among researchers due to its ability to deal with uncertain, imprecise and incomplete data. This paper develops a fuzzy logic-based model for predicting the maintainability of a class. The given model is based on the Mamdani’s fuzzy inference system. Chidamber and Kemerer metrics are used as inputs to the model and the maintainability is computed as output. Maintainability can be used as an indicator of the quality of software at design time.