Artificial Intelligence in Software Test Automation: A Systematic Literature Review

Artificial intelligence (AI) has made a considerable impact on the software engineering field, and the area of software testing is not an exception. In theory, AI techniques could help to achieve the highest possible level of software test automation. The goal of this Systematic Literature Review (SLR) paper is to highlight the role of artificial intelligence in the software test automation area through cataloguing AI techniques and related software testing activities to which the techniques can be applied. Specifically, the potential influence of AI on those activities was explored. To this end, the SLR was performed with the focus on research studies reporting the implementation of AI techniques in software test automation. Out of 34 primary studies that were included in the final set, 9 distinct software testing activities were identified. These activities had been reportedly improved by applying the AI techniques mostly from the machine learning and computer vision fields. According to the reviewed primary studies, the improvement was achieved in terms of reusability of test cases, manual effort reduction, improved coverage, improved fault and vulnerability detection. Several publicly accessible AI-enhanced tools for software test automation were discovered during the review as well. Their short summary is presented.