Using a class abstraction technique to predict faults in OO classes

Abstract

In this paper, we propose an innovative suite of metrics based on a class abstraction that uses a taxonomy for OO classes (CAT) to capture aspects of software complexity through combinations of class characteristics. We empirically validate their ability to predict fault prone classes using fault data for six versions of the Java-based open-source Eclipse Integrated Development Environment. We conclude that this proposed CAT metric suite, even though it treats classes in groups rather than individually, is as effective as the traditional Chidamber and Kemerer metrics in identifying fault-prone classes.

Automatic patch generation learned from human-written patches

“…We used Eclipse JDT, because it has a long revision history (more than 10 years), and is widely used in the literature [11], [12]. We used the Kenyon framework [13] to retrieve bug patches [14].…”

Adaptive Software Fault Prediction Approach Using Object-Oriented Metrics

“…Following the selection process, our adaptation approach utilizes general transformation function on both sets of metrics, training 4 As measured by F-Score, the harmonic mean of precision and recall defined in Section 4.6…”

“…While CBO turned out to be an overall good fit for our fault prediction, in this dissertation, we solely consider product metrics. We plan to expand the failure data investigation to proprietary software systems due to the different development practices used during their development, and particularly by including more sophisticated product measures [4], as well as process and resource metrics to complement our product measures [59].…”