Software requirements (SRs) modeling is one of the sub-processes of requirements engineering whose aim is to model and design the SRs before the development of a project. Different techniques are employed for representing the SRs using goal concepts, unified modeling language, etc. There could be a large number of SRs after the completion of the requirements elicitation process. It is not practical to model the complete set of the identified SRs because of the cost, time, and other limitations of an organization. There should be some systematic methodology to identify and select those SRs for modeling that need to be implemented during the software development process. The selection of SRs from the list of the elicited requirements is a multi-criteria decision-making process. In this process different stakeholders participate in the selection of the SRs. Linguistic variables may be used by the stakeholders to specify the preferences of SRs. To deal with this issue, a method has been proposed using a fuzzy based approach so that the selected set of SRs can be modeled and implemented during the development phase. The proposed method is explained by considering the small and large set of SRs for an institute examination system. The ranking value of the functional requirements of an examination system is computed. Based on the ranking order, top three requirements are modeled using use-case diagrams (UCDs) and class diagrams. It is found that both diagrams represent different information about the requirements of an examination system and there is no overlap in the information captured through UCDs and class diagrams.