Usability Assessment of the Use-Case Model Textual Specification Language

The requirements specification is a key outcome of requirements engineering. It serves as input to all phases of software development, including design, implementation, and testing. On one hand, due to the diverse competencies of stakeholders, it should be written in a clear and understandable language, preferably natural. On the other hand, it should be structured and computationally processable. Controlled languages, with limited syntax, partially meet both demands, albeit restricting the freedom of requirement expression. This raises the question of whether this limitation affects the overall usability of the specification language from the perspective of its users. To address this question, a controlled experiment was conducted to assess the usability of the Use-Case Model Textual Specification Language (T-UFCL). Its usability was compared with alternative notations – informal (natural language) and formal (activity diagrams). The experiment results showed that the controlled language’s usability is comparable to natural language and, in certain cases, higher than that of formal notations.