What is it about?
This paper presents an automated approach for generating sequence diagrams from user stories. Additionally, it explores the effectiveness of Large Language Models (LLMs), particularly ChatGPT, in performing the same task. The study compares the sequence diagrams produced by the proposed automated method with those generated by ChatGPT, providing insights into the accuracy, efficiency, and potential of LLMs in software modeling tasks.
Featured Image
Photo by Igor Omilaev on Unsplash
Why is it important?
This paper presents an automated method to generate sequence diagrams from user stories, significantly improving efficiency and accuracy in software modeling. By comparing this approach with ChatGPT's performance in generating sequence diagrams, it evaluates the potential of large language models (LLMs) in software engineering tasks. The findings offer insights into the benefits of automation and AI integration in streamlining development processes, while also revealing areas for future improvements in model-driven development.
Perspectives
Working on this paper was an exciting experience, particularly in exploring the potential of ChatGPT to assist software engineers in model-driven development. It was fascinating to witness how a large language model could be leveraged to automate sequence diagram generation from user stories, a traditionally manual task. The results not only opened new avenues for integrating AI into software development workflows but also highlighted areas for future refinement, making this research a valuable step towards more efficient and AI-supported engineering practices.
Munima Jahan
Thompson Rivers University
Read the Original
This page is a summary of: Automated Derivation of UML Sequence Diagrams from User Stories: Unleashing the Power of Generative AI vs. a Rule-Based Approach, September 2024, ACM (Association for Computing Machinery),
DOI: 10.1145/3640310.3674081.
You can read the full text:
Resources
Contributors
The following have contributed to this page
