Leveraging Generative AI and Machine Learning in Software Testing: Emerging Tools and Technologies for Quality Engineering
R. A. Alagu Raja · International Scientific Journal of Engineering and Management · 2023
Introduction: Software testing, a vital SDLC phase, ensures software meets functional, performance, and security standards. Generative AI and ML are transforming testing by automating tasks like test case generation, execution, and analysis, enhancing efficiency and flaw detection. Leveraging AI-driven tools enables superior quality, broader test coverage, and intelligent decision-making, revolutionizing the software testing landscape and driving high- quality software delivery. Methods: This study uses a systematic review methodology, adhering to PRISMA guidelines, to explore AI and ML integration in software testing. It examines trends, tools, and gaps, addressing research questions on manual testing drawbacks, AI's role, automation tasks, and assessment techniques. A rigorous protocol ensures comprehensive, reliable, and reproducible findings. For this study focused on articles published between 2016 and 2023 across six databases, using a comprehensive search string. After screening 90 articles, 20 studies were included. Data extraction emphasized AI techniques, testing tasks, and performance metrics. Results: The review highlights advancements in ML and DL for software testing from 2016 to 2022. Early studies focused on manual testing and foundational ML techniques, while later years emphasized automation, GUI testing, and fault prediction using deep learning. Tools like ARES, DeepOrder, and DeepXplore showcase innovations in prioritization, coverage, and efficiency, reflecting the evolving scope of ML/DL in software testing. Conclusion: This review showed a greater role of Generative AI and Machine Learning in software testing and quality engineering, offering innovative solutions for automation and accuracy. Keywords: Software Testing, Machine Learning, Deep Learning, Automated Testing and AI in Testing