Assessing Weight of the Coupling between Objects towards defect forecasting in Object Oriented Programming
This study presents a new approach for predicting fault proneness in object classes. A new metric called weighted coupling between objects or wcbo, measures a new unit called weighted coupling between objects. The approach then ranks each class of the object based on its fault proneness using the HITS algorithm that measures hub and authority weights in link analysis. The proposed metrics advantage lies in the fact that it can be computed without using any programming language, in a simpler way, comparatively. The study undertook with huge open source systems to forecast fault proneness, shows the projected wcbo fault proneness sensitivity to be 98.7% and compared with CBO which was 42% sensitive.
