Quantitative Indicator for Objective Assessment of Software Process Quality and Performance

Abstract

A company's software production and product quality can be improved by evaluating the software development process. Artefact inspection is one example of a flawed traditional method that relies on manual qualitative evaluations and is time-consuming, constrained by authority, and frequently subjective. To address these limitations, this study introduces an innovative, semi-automatic method for assessing software processes, leveraging machine learning techniques. We define the problem as a sequence classification challenge that can be effectively tackled using machine learning algorithms. Building on this framework, we develop a new quantitative metric for the objective evaluation of software process efficiency and quality. To validate the effectiveness of our approach, we apply it to evaluate the defect management procedures employed in four real-world industrial software projects. Our empirical findings demonstrate that our method is effective and has the potential to deliver reliable, quantitative evaluations of software processes.

Keywords

• Machine Learning
• Software Processes
• Quantitative Evaluation
• Software Process Quality