Change classification, today known as Just-in-Time Defect Prediction, is a technique for predicting software bugs at the change level of granularity. Several ideas came together to form change classification: predictions on code changes, using word-level textual features, use of machine learning classifiers, and leveraging open source code repositories. While change classification has led to a robust line of research, it has not yet had significant industrial adoption. A key recommendation is to explore explainability features so developers can better understand why a prediction is being made. We explore how large language models can advance this work by providing prediction explanations and bug fix suggestions.

Machine learning classifiers have recently emerged as a way to predict the introduction of bugs in changes made to source code files. The classifier is first trained on software history, and then used to predict if an impending change causes a bug. Drawbacks of existing classifier-based bug prediction techniques are insufficient performance for practical use and slow prediction times due to a large number of machine learned features. This paper investigates multiple feature selection techniques that are generally applicable to classification-based bug prediction methods. The techniques discard less important features until optimal classification performance is reached. The total number of features used for training is substantially reduced, often to less than 10 percent of the original. The performance of Naive Bayes and Support Vector Machine (SVM) classifiers when using this technique is characterized on 11 software projects. Naive Bayes using feature selection provides significant improvement in buggy F-measure (21 percent improvement) over prior change classification bug prediction results (by the second and fourth authors [28]). The SVM's improvement in buggy F-measure is 9 percent. Interestingly, an analysis of performance for varying numbers of features shows that strong performance is achieved at even 1 percent of the original number of features.

Research on Mining Software Repositories (MSR) has yielded fruitful results in many Software Engineering areas including software change comprehension, bug prediction, and developer network recovery. When performing MSR research, the first task is to extract features corresponding to source code details from repositories. Since reusable feature extraction tools are not available, each MSR research group builds their own extraction tool, a duplication of effort. We introduce a reusable feature extractor, Kenyon-web, for MSR research. Kenyon-web is fully reconfigurable, pluggable, and serves most MSR related tasks. In this report, we show the architecture of Kenyon-web and demonstrate its utility by showcasing a sample MSR task.

Recently, machine learning classifiers have emerged as a way to predict the existence of a bug in a change made to a source code file. The classifier is first trained on software history data, and then used to predict bugs. Two drawbacks of existing classifier-based bug prediction are potentially insufficient accuracy for practical use, and use of a large number of features. These large numbers of features adversely impact scalability and accuracy of the approach. This paper proposes a feature selection technique applicable to classification-based bug prediction. This technique is applied to predict bugs in software changes, and performance of Naive Bayes and Support Vector Machine (SVM) classifiers is characterized.