An Integrated Analytical Approach to Mathematical Modelling the Stochastic Behaviour of Software Debugging Phenomenon
lMohd Taib' Shatnawi, Omar Shatnawi · 2021
Analytical and stochastic modelling methods can provide quantitative measures of software systems' reliability during the debugging process. As software increases in complexity, debugging assumes more development time and becomes less deterministic. Many of the factors influencing the debugging-process activities are stochastic, such as effort, complexity, efficiency, and debugger skill. With this motivation, in this paper, an integrated software-reliability modelling approach based on a Poisson process to address these factors subject to the categorization of the faults under imperfect-debugging and learning-process phenomena. The software faults encountered are categorized into three types based on their debugging-complexity. The debugging-complexity is proportional to the level of effort expenditures required to correct the fault. Correspondingly, the entire debugging-process is the summing of the three debugging activities processes. A numerical example is provided based on a real software-reliability dataset to demonstrate the descriptive and predictive power capabilities and evaluate software-reliability measures as per the proposed integrated modelling approach and other well-documented models.