Object-Oriented Design Metrics

Abstract

The thesis is organized as follows: Chapter 1 includes an introduction to the research done in the thesis. Chapter 2 explores related work from the literature concerning different aspects, including object-oriented software metrics, software defect prediction, and different data preprocessing techniques used to enhance the defect prediction performance. The software defect prediction section in¬cludes both within-project defect prediction and cross-project defect pre¬diction. Chapter 3 reviews the object-oriented software metrics used by the proposed prediction models, the artificial neural network algorithm used to build the proposed model, and the quantiles that are used in the preprocessing of different metrics. Chapter 4 delves into the proposed universal defect prediction model, including details about how datasets have been collected and how met¬rics are measured. It also includes the explanation of different data preprocessing approaches applied to different metrics sets and details about the construction of the universal defect prediction model. Chapter 5 defines the different metrics used to evaluate the model’s performance. Chapter 6 presents and discusses the results of this study. Chapter 7 considers the threats to the validity of this study. Finally, Chapter 8 concludes this research and points to potential future work.