Prioritising scenario evolution

Use cases are currently used in requirement elicitation and analysis. They describe in natural language the complete functionality of a proposed system. In most projects, the amount of detail put into eliciting and documenting a use case is usually dependent on the time and resources available. Seldom is there enough time to fully expand all the scenarios in the use cases. Thus, only a small number of the scenarios are ever fully expanded. This paper delves into the process of expanding the scenarios of use cases. The transition from use cases to OO modelling elements requires functionally decomposed use cases. Discovering possible scenarios reveals the system facilities that are needed. They also guide the discovery of all the relevant classes and objects, and exposes the range of possible system interactions in the solution domain. Thus, it is important to discover and expand the most important scenarios in the use cases. Each scenario of the high level use cases can be ranked by applying various measures from the use case model, from the stakeholders´ goals, and from applying concepts and formulas from Social Network Analysis. Once the most important scenarios are detected, the elaboration of a use case can be properly directed and controlled. The technique used for ranking scenarios has been applied to the case study of developing a meeting scheduler system for JRCASE, Macquarie University. The findings as well as observations on reusability and dependencies in the use case model are presented and analysed