Artificial intelligence in software and service lifecycle

Abstract

Artificial Intelligence is an emerging technology with a huge potential to revolutionize the Software Development in the IT industry. As applications grow in complexity and scale, maintenance is becoming a challenge. Leveraging AI in automating maintenance tasks such as bug fixing, detecting code smells, managing technical debts, and code refactoring ensures the longevity and optimal performance and efficient usage of resources. This thesis provides a systematic mapping study with analysis of 121 research articles in the domain of AI in IT industry at all the phases of SDLC right from Planning till Maintenance with a special focus on the maintenance activities. It provides a demonstration of a prototype using deep learning Recurrent neural network detecting a code smell (a maintenance activity). The findings of the systematic mapping study provide the application and impact of AI in various developmental and maintenance activities. The results of the demonstration show practical implementation of research study. It also identified research gaps in a few areas, suggesting a need for further research and development in those areas.