Application of AI for Automation of Software Testing
International Association for Computer and Information Science
Faculty of Computing, Health and Science
School of Computer and Security Science
Software testing forms an essential role in the development and distribution of any software application. It is essentially a human centred process which requires high levels of expertise from the software engineers involved. The cost, time and effort involved in exhaustively testing any software application is prohibitive and impractical. The automation (or semi-automation) of this process would go a long way to reducing the cost of the software testing process. If these processes were robust then the quality of the resulting software would also be improved. From its early beginnings, the role of AI has been to develop methodologies which define, refine and explore large search spaces. This approach is ready made for its application to the problem of software testing. Existing work in the application of Al techniques to software testing has been restricted to toy problems raising scalability issues. A more in depth approach is required. Experts in software engineering need to collaborate with experts in artificial intelligence so that existing AI methodologies can be better employed in software testing. This paper hopes to serve as a stimulus for such cross pollination.