Automated Derivation of UML Sequence Diagrams from User Stories: Unleashing the Power of Generative AI vs. a Rule-Based Approach
Munima Jahan, Mohammad Mahdi Hassan, R. M. Hashemi Golpayegani, Golshid Ranjbaran, Chanchal K. Roy, Banani Roy, Kevin A. Schneider
Abstract
User stories are informal, non-technical descriptions of features from a user's perspective that guide collaboration and iterative development in Agile projects. However, ambiguities in user stories can lead to miscommunication among stakeholders. Design models, such as UML sequence diagrams, are essential for enhancing communication, clarifying system behavior, and improving the development process. This paper presents an automated approach for generating behavioral models specifically sequence diagrams from natural language requirements expressed as user stories. We also investigate the effectiveness of a Large Language Model (LLM) in using generative AI for this task. By applying our approach and ChatGPT to two benchmark datasets with the same set of user stories, we generated corresponding sequence diagrams for comparison. Expert evaluations in Software Engineering reveal that our approach effectively produces relevant, simplified diagrams for straightforward user stories, whereas the LLM tends to create more complex diagrams that sometimes go beyond the simplicity of the original user stories.