Towards a strategy for software requirements selection
Abstract: The importance of identifying clearly the requirements for a software system is now widely recognized in the software engineering community. From the emerging field of requirements engineering, this thesis identifies a number of key areas. In particular, it shows that, to achieve customer satisfaction, it is essential to select a subset of all candidate requirements for actual implementation. This selection process must take into account both the value and the estimated cost of including any candidate requirements in the set to be implemented. At the moment this entire process is usually carried out in an informal manner so there is a pressing need for techniques and a strategy to support it.With the explicit aim of clearly identifying requirements, the use of quality function deployment (QFD) in an industrial project at Ericsson Radio Systems AB was studied and evaluated. It was found that QFD helps developers to focus more clearly on the customer’s needs and in managing non-functional requirements.To trade off maximization of value and minimization of cost when selecting requirements for implementation a method was developed, the contribution-based method, which is based on the analytic hierarchy process (AHP). The contribution-based method was applied in two industrial projects and the results evaluated. Both studies indicate that both the value and cost of candidate requirements can vary by orders of magnitude. Thus, deciding which requirements to be selected for implementation is of paramount importance, and also a primary determinant of customer satisfaction. The contribution-based method forms the basis of a selection strategy that will maximize customer satisfaction.
This dissertation MIGHT be available in PDF-format. Check this page to see if it is available for download.