Expert Systems with Applications vol:39 issue:9 pages:8115-8135
In this paper, we propose an expert system for iterative requirements engineering using Formal Concept Analysis. The requirements engineering approach is grounded in the theoretical framework of C–K theory. An essential result of this approach is that we obtain normalized class models. Compared to traditional UML class models, these normalized models are free of ambiguities such as many-to-many, optional-to-optional or reflexive associations which cause amongst others problems at design time. FCA has the benefit of providing a partial ordering of the objects in the conceptual model based on the use cases in which they participate. The four operators of the C–K design square give a clear structure to the requirements engineering process: elaboration, verification, modification and validation. In each of these steps the FCA lattice visualization plays a pivotal role. We empirically show how the strategy works by applying it to a set of case studies and a modeling experiment in which 20 students took part.