International Journal of Computer Applications (0975 – 8887) Volume 94 – No 9, May 2014 17 Fuzzy Logic Approach to Forecast the Extendibility/Extensibility in Object Oriented Design using an Integrated Model Rajinder Vir A.P CTIEMT, Shahpur Jalandhar, Punjab, India Parwinder Dhillon A.P APEEJAY,Jalandhar Punjab, India Jaswinder Dhillon A.P DAVIET, Jalandhar Punjab, India ABSTRACT A number of researchers have conducted various empirical studies on the software metrics for Object Oriented design. The research proved that some of these metrices are very useful for forcasting the quality attributes of the software like extendibility/extensibility, effectiveness, reliability and maintainability. In this paper a hybrid approach is proposed for investigating the extendibility/extensibility of classes in Object Oriented design.Tthe hybrid approach will comprised of subset of CK netric suite and mood netric suite. These days a great demand occur for finding software measurement so that quality of software can be forcasted. Therefore software engineering require various quality models that can be used for forcasting the characteristics for quality such as extendibility/extensibility, effectiveness, reliability and maintainability. The main objective of this work is to experimentally forecast the association between OOD metrics and extendibility/extensibility. General Terms Software Engineering, object Oriented design, extendibility, extensibility, Classes in object oriented Design. Keywords MOOD Metric suite, CK metric suite, Fuzzy inference system, Mamdani inference model. 1. INTRODUCTION Object Oriented Programming is a programming paradigm that represents concepts as objects and accompliced procedures known as methods. Objects are used so that applications and computer programs can be easily be interacted. OOP is essential for software developments, because it determines the structure of the software solution in an appropriate manner. Once the design is prepared, it is difficult to apply modifications and also it becomes very expensive. Therefore design should be focused from the beginning. Software metrics are the measurements that can be defined to forecast the quality of the software during the early phases of software development process. Metrics can be used to figure out the design quality. Many Metrics have been proposed for OOP. It is analyzed from the previous research that CK metrics suite [13] and the MOOD metric suite [3] are found to be best to calculate the OOP quality. It is analyzed that software developers need combination of metrics to predict the quality of software. Hence there should be some integrated approach to combine these software metrics into a single output unit. Metrics offer a mechanism for attaining more accurate estimation of project milestones, and developing a software system that contains minimum faults [1]. There are a number of OOP software metrics available these days. These metrics are very helpful in fetching the information about the quality of OOP software. In this paper the way has been described how the evaluation of the extendibility/ extensibility using the software metrics has been done - CK metrics viz DIT and MOOD metrics viz MHF, AHF, AIF and MIF. The input metrics are divided into three linguistic terms low, medium and high. The paper will proceed as follows: Section II will present the literature survey. Software quality along with its characteristics will be presented in section III. Introduction to software metrics will be presented in section IV. In next section V, paper will present the proposed integrated model based on fuzzy logic. In section VI paper will describe the evaluated experimental results. Finally in section VII, paper will discuss the related conclusion and the future scope. 2. LITERATURE REVIEW A large number of metrics have been proposed in the past for so many years to confine the OO design, code and constructs. These metrics provide ways to assess the quality of software and their use in early phases of software development which can help software companies in evaluating large software development quickly and at a reasonable cost [4]. There have been large number empirical studies evaluating the impact of OO metrics on faulty classes. Saxena et al. [6] provided a review of all those empirical studies from 1995 to 2010 to predict software fault proneness with a specific focus on techniques used. Benlarbi et al. [11] surveyed that the basic premise behind the development of object oriented metrics is that they can serve as early predictors of classes that contain faults or that are closely maintain. They have shown that size can have an important confounding effect on the validity of object-oriented metrics. Khalsa [16] proposed an algorithm using fuzzy logic to measure fault proneness and defect density of the software development process and hence can be used to minimize rework. Kamiya et. al. [5] proposed a new method to estimate the fault-proneness of an object class in the early phase, using several complexity metrics for object-oriented software. Four checkpoints were introduced into the analysis/ design/ implementation phase, and estimates were done on the fault- prone classes using applicable metrics at each checkpoint. Menzies et al. [14] compared Decision Trees, Naïve Bayes,