Data-Driven Software Engineering: The Integral Role Of Generative Ai In Transforming Software Development Processes

Abstract

This thesis examines the transformative convergence of data science and software engineering, with a particular emphasis on how generative AI technologies are revolutionizing traditional software development processes. The thesis introduces a framework for the ethical and effective implementation of generative AI across the software development lifecycle. There has been a significant impact of emerging technologies on software engineering processes over the past few years. Consequently, there has been a shift from experience based to data-driven decision-making. This transition towards data-driven decision making has led to more precise and reliable decisions, resulting in more effective and efficient software engineering processes with reduced rework. Our study consists of two stages. The first stage includes a comprehensive systematic literature review on the applications of data-driven methodologies in software engineering processes from the past decade. Through the analysis of 34 primary studies, we identified the domains where the use of data-driven methodologies is most prevalent. The analysis showed that data driven approaches are widely applied in software management and software testing within software engineering processes. Researchers have been utilizing various subfields of artificial intelligence, including machine learning and deep learning, to develop decision models for software engineering processes. In the second section of our study, to satisfy a gap determined as a result of the SLR, we proposed a framework for utilizing generative AI in software engineering processes to reduce costs, shorten project timelines, improve workflow efficiency, and enhance software quality. The framework outlines how data should be processed and how generative AI can be leveraged in software engineering processes, as outlined in SWEBOK v4, and presents the benefits and limitations of its application.