Some Recent Work in Computational Intelligence for Software Engineering

Computational intelligence has been used in software engineering for a long time. There has been a recent surge in interest in this area, especially in search-based software engineering. This talk touches upon some of the recent examples in the broader field of computational intelligence in software engineering. It is highlighted that software engineering could benefit from advanced computational intelligence techniques in tackling hard problems, e.g. software module clustering, software reliability maximisation, software project scheduling, software effort estimation, software defect prediction, etc. It is also argued that new research challenges posed by software engineering could stimulate further development of new theories and algorithms in computational intelligence. Such theoretical research could shed some light on important research issues and provide guidance in future work. For example, theoretical analysis of computational time complexity of search algorithms can inform us about the limitation of search-based software engineering. The research in online learning algorithms can help us develop novel approaches to software effort estimation when historical data within a company are sparse. The primary aim of this talk is not to provide a comprehensive review of computational intelligence for software engineering, but to illustrate the opportunities for further research and development in this area through selected examples.