Change prediction model using evolutionary history of a software — An automated tool
Ruchika Malhotra, Aryan Aryan, Aastha Gupta · 2017 4th IEEE Uttar Pradesh Section International Conference on Electrical, Computer and Electronics (UPCON) · 2017
The Software systems tend to undergo a lot of changes as we go from one version to another. In order to accurately measure the change proneness of the software we developed tool that automatically calculates the change history of the classes. The change history is in the form of Evolution Based metrics and the Object Oriented metrics are also used. We calculated the accuracy in calculating the dependent variable of the research i.e. change proneness of the class using four different prediction models. The results of the paper show that the prediction model including Evolution Based metrics and the Object Oriented metrics provides the most accurate predictions of the change proneness. The tool initially takes different versions of the software and finally calculates the accuracies of different prediction models.