Type-2 fuzzy logic based prediction model of object oriented software maintainability

Abstract

In this work, a maintainability prediction model for an object-oriented software system based on type-2 fuzzy logic system is presented. With the proliferation of object-oriented software systems, it has become very essential for concerned organizations to maintain those systems appropriately and effectively. However, it is pathetic to note that just very few number of maintainability prediction models are currently available for object oriented software systems. In this work, maintainability prediction model based on type-2 fuzzy logic systems is developed for an object-oriented software system. Earlier published objectoriented metric dataset was used in building the proposed model. Comparative studies involving the prediction accuracy of the proposed model was carried out in relation to the earlier used models on the same datasets. Empirical results from experiments carried out indicates that the proposed type-2 fuzzy logic system produced better and interesting results in terms of prediction accuracy measures authorized in object oriented software maintainability literatures. In fact, the proposed method satisfies the three major conditions stated in the literatures as basis to determining a good maintainability prediction model.