Home // IARIA Congress 2025, The 2025 IARIA Annual Congress on Frontiers in Science, Technology, Services, and Applications // View article
Authors:
Maxmillan Giyane
Dzinaishe Mpini
Keywords: Use Case Diagram; Large Language Model; GPT 3.5 Turbo; Natural Language Processing.
Abstract:
Software Architects often use Use Case diagrams, a type of Unified Modelling Language (UML) behavior diagram, to capture user needs and system functionalities. These diagrams aid in project estimation by identifying system requirements and reducing ambiguity. Creating them manually is a time-consuming task prone to errors. This research aims to automate Use Case diagram generation from text using the Generative Pretrained Transformer 3.5 (GPT-3.5) Turbo model. The developed tool uses a Natural Language Processing (NLP) technique to extract actors, use cases, and associations from descriptions, and convert these elements into UML-compliant diagrams. It also includes an interactive interface for Use Case diagram refinement. The system processes user input text to identify relevant elements, visualizes them using jCanvas, and allows real-time user interaction for refinement. Testing showed an 89.33% accuracy in element identification but highlighted areas for improvement like handling edge cases and optimizing performance. This research demonstrates the potential of NLP and visualization tools to improve Use Case diagram generation efficiency and accuracy, with future work focusing on enhancing usability and functionality.
Pages: 47 to 51
Copyright: Copyright (c) IARIA, 2025
Publication date: July 6, 2025
Published in: conference
ISBN: 978-1-68558-284-5
Location: Venice, Italy
Dates: from July 6, 2025 to July 10, 2025