A FUZZY LOGIC-BASED METHODOLOGY FOR THE ACQUISITION AND ANALYSIS OF IMPRECISE REQUIREMENTS

Two major challenges with requirement analysis in concurrent engineeri ng are: (1) requirements from multiple members of a concurrent enginee ring team are often conflicting with each other; and (2) requirements are often imprecise in nature. Existing formal methods for requirement engineering are very limited in addressing these issues. More specifi cally, they have not fully explored the use of artificial intelligence technique for achieving effective trade-offs among conflicting imprec ise requirements. This paper presents a comprehensive methodology for specifying imprecise requirements and for characterizing complex relat ionships among them to facilitate trade-off analysis. Imprecise requir ements are represented by the canonical form in test-score semantics i n fuzzy logic. A formal approach and a practical method are developed to analyze the complex relationships between requirements. Conflicting requirements can be identified and represented using both qualitative terms and quantitative measures. Multiple requirements with complex r elationships among them are fused into an overall system requirement b ased on fuzzy multi-criteria decision techniques. To obtain a feasible overall system requirement that is satisfactory to customers, the ite rative refinement of requirements and the negotiation between the cust omers and the requirement analysts regarding conflicting requirements are crucial. Our methodology supports the iterative process of refinem ent end negotiation by facilitating a formal trade-off analysis, by pr oviding intelligent feedbacks generated based on the analysis, and by defining a clear process of compromise. Therefore, this methodology ca n help to achieve a better system objective that is satisfactory to cu stomers and feasible to developers by fully exploiting the elasticity of imprecise requirements. In addition, the explicit specification of imprecise requirements provides a basis for verification and validatio n of software systems.