Improving Software Defect Prediction by Aggregated Change Metrics

To ensure the delivery of high quality software, it is necessary to that all its artifacts function properly, which usually done by performing appropriate tests with limited resources. It therefore desirable identify defective so they can be corrected before testing process. So far, researchers have proposed various predictive models for this purpose. Such are typically trained on data representing previous project versions a software and then used predict in new version likely defective. However, consists measurable properties or modules, leaves out information about timeline development fill gap, we propose set metrics, namely aggregated change created aggregating changes made between two versions, taking into account chronological order changes. In experiments conducted open source projects written Java, show stability performance commonly classification improved extending feature include both analyzed metrics.