Automating the reuse of domain knowledge to improve the modelling outcome from interactions between developers and clients
Informal interactions with clients are prone to errors in interpretation and communication which can later result in errors in the ensuing requirements models. Resolving such errors early can prevent rework or the building of a system that is non-compliant with the client's intention. In developing complex systems and highly interactive systems, such as agent oriented systems, those errors in requirement analysis are even more likely. By focussing on the quality of interactions between the future users of the system and the human modellers, this article presents an innovative validation and verification process to support analysis of requirement analysis. We use domain knowledge to validate compliance with the client's intention. Using layered abstractions, we formulate the validation process which is methodology independent. An interactive support tool is implemented to support the developers using the requirement analysis process. The process is conducted iteratively to accommodate the lifecycle defined by most methodologies. The process and the tool are validated and illustrated using two different development exemplars illustrating different domains and different development methodologies.