Optimizing Cost Efficiency in Software Defect Prediction Through Network Representation Methods
Keywords:
Software Defect Prediction, Generative Adversarial Networks, Network Representation, Class Dependency Networks, Call Graphs, Cost AnalysisAbstract
The significance of software defect prediction (SDP) has been well established owing to its usefulness in preventing potential defects in software at the earliest possible phase within its development cycle. Research works in SDP utilizing traditional metrics related to code complexity and coupling does not have the capability to capture the interrelationships and interactions that are a common characteristic in big software systems. A better modelling of the underlying structural relationships within the software is required to design an efficient and accurate SDP model. Network based graphical representations such as call graphs and class dependency
networks have the potential to capture the intracies of the dependencies and the hidden patterns among those dependencies.
References
Li, Z., Niu, J., & Jing, X. Y. (2024). Software defect prediction: future directions and challenges. Automated Software Engineering, 31(1), 19, https://doi.org/10.1007/s10515-024-00424-1
Wu, W., Wang, S., Liu, B., Shao, Y., & Xie, W. (2024). A novel software defect prediction approach via weighted classification based on association rule mining. Engineering Applications of Artificial Intelligence, 129, 107622, https://doi.org/10.1016/j.engappai.2023.107622
Boucher, A., & Badri, M. (2018). Software metrics thresholds calculation techniques to predict fault proneness: An empirical comparison. Information and Software Technology, 96, 38-67, https://doi.org/10.1016/j.infsof.2017.11.005
