| Abstract [eng] |
Requirements specification and analysis is highly dependent on the analyst’s knowledge and skills, which in turn increases the empirical factor in requirements specification. Therefore, in the process of user requirements specification it is expedient to use the knowledge base created on the basis of enterprise metamodel, which storages the subject area knowledge (enterprise model), thus increasing the efficiency of requirements specification process by reusing enterprise models knowledge. This will not only reduce the empirical factor and the likelihood of errors, but also save analysts time. In order to implement the method of creating a requirements specification, an analysis had to be performed: software that is designed to specify requirements, requirements specification templates and knowledge-based systems and their business model. Then the proposed method was specified and an experimental system was designed to test the performance of the method with real data. In the study, the requirements description forms used in the method were compared with the composition of the user requirements specification standards and it was found that the Volere template could be covered up to 97% of the content of the user requirements specification document. Also, using as many fields as possible in the description forms with limited input values and choices can reduce the empirical factor of requirements specification. |
