Optimising software development policies for evolutionary system requirements

Anticipating future software requirements might support the evolution of software systems and as such reduce the cost of development and maintenance in due time. Unfortunately identifying the right set of evolution scenarios is difficult due to the uncertainty of occurrence of future requirements. In this paper we propose the Software Evolution Analysis Model (SEAM) that provides a probabilistic model for evolution requirements, which can more precisely anticipate future requirements and as such reduce the chance on unanticipated requirements. In SEAM the cost of each individual evolution requirement is defined and an optimal set of evolution scenarios is calculated using dynamic programming techniques.