Construction of Membership Function for Software Metrics

Abstract

A software developer has to deal with a lot of challenging requirements such as cost prediction, defect prediction, reliability prediction, testing effort prediction, safety prediction, and many more while developing quality software. However, it has been found that the most of the software development activity is performed by human beings. This may introduce various faults across the development, causing failures in near future. Therefore, prediction of software defect has been one of the major areas of concern. A number of the software defect prediction model using software metrics has been proposed in last two decades. However, predicting software defect by taking all the software metrics (traditional, object oriented and process) is computationally complex. Therefore, an intelligent selection of metrics plays a vital role in improving the software quality. In the early phases of the software development life cycle, software metrics are associated with uncertainty and can be assessed in linguistic terms. Construction of membership function is very important because the success of a method depends on the membership functions used. Therefore, in this paper, a methodology has been proposed to construct the membership functions of software metrics.