Requirements elicitation step is of paramount importance in the requirements engineering process. In the distributed environment of so-called inter-company cooperative information system, this step is a thorny issue. ...Requirements elicitation step is of paramount importance in the requirements engineering process. In the distributed environment of so-called inter-company cooperative information system, this step is a thorny issue. To elicit require-ments for an inter-company cooperative information system, we early proposed a methodology called MAMIE (from MAcro to MIcro level requirements Elicitation) with an accompanied tool. In MAMIE methodology, requirements are the result of composing functional and non-functional concerns. Before non-functional concerns composition, it’s primary to identify relationships between them. According to the most existing approaches, a non-functional concern may have a negative, positive or null contribution on the other non-functional concerns. In this paper, we argue that using only these three contributions types is not sufficient to express relationships which may exist between non-functional concerns. Thus, we propose a process which aims to identify non-functional concerns’ relationships and model them using a fuzzy cognitive map. The resulting model is composed of non-functional concerns, relationships between them and the weight of these relationships expressed with linguistics fuzzy values. Using fuzzy cognitive maps to model non-functional concerns relationships allows moving from the conventional modelling toward developing a computer based model. An example from the textile industry is used to illustrate the applicability of our process.展开更多
文摘Requirements elicitation step is of paramount importance in the requirements engineering process. In the distributed environment of so-called inter-company cooperative information system, this step is a thorny issue. To elicit require-ments for an inter-company cooperative information system, we early proposed a methodology called MAMIE (from MAcro to MIcro level requirements Elicitation) with an accompanied tool. In MAMIE methodology, requirements are the result of composing functional and non-functional concerns. Before non-functional concerns composition, it’s primary to identify relationships between them. According to the most existing approaches, a non-functional concern may have a negative, positive or null contribution on the other non-functional concerns. In this paper, we argue that using only these three contributions types is not sufficient to express relationships which may exist between non-functional concerns. Thus, we propose a process which aims to identify non-functional concerns’ relationships and model them using a fuzzy cognitive map. The resulting model is composed of non-functional concerns, relationships between them and the weight of these relationships expressed with linguistics fuzzy values. Using fuzzy cognitive maps to model non-functional concerns relationships allows moving from the conventional modelling toward developing a computer based model. An example from the textile industry is used to illustrate the applicability of our process.
