A formal methodology for analyzing tradeoffs of imprecise requirements

Conflict identification and resolution are inevitable parts of the requirement analysis process. Over the past few years, the need to deal with conflicting system requirements has become increasingly important. These requirements are often elastic in that they can be satisfied to a degree. The overall goal of this research is to develop a formal methodology that facilitates the identification and tradeoff analysis of conflicting requirements by explicitly capturing their elasticity. In order to capture the elasticity of imprecise requirements, we represent imprecise requirements using fuzzy logic. Based on the representation, we build a formal foundation to facilitate the identification of conflicting requirements. Once the conflicting requirements are identified, we describe a systematic approach for analyzing the tradeoff between conflicting requirements using the techniques in decision science. The systematic tradeoff analyses are used for three important tasks in the requirement engineering process: (1) for validating the structure used in aggregating prioritized requirements, (2) for assisting requirement engineers in identifying the structures and the parameters of the underlying representation of imprecise requirements and in eliciting them from the customer, and (3) for assessing the priorities of conflicting requirements. We illustrate the usage of these techniques using the requirements of a conference room scheduling system.