A new approach for automatic test case generation from use case diagram using LLMs and prompt engineering

Abstract

The automation of test case generation from UML diagrams is a growing field that aims to make the software development process smoother. This paper suggests a new framework that uses generative artificial intelligence (AI) to turn use case diagrams into test cases that can be executed. By getting information from the XML representation of use case diagrams, we can create detailed instructions that guide a generative AI model to make test cases for each use case scenario. This method not only makes test case creation easier but also ensures we cover everything well and accurately, which could make software products get to market faster. this approach shows how traditional software engineering methods and new AI techniques can work well together, giving us an idea of what automated software testing might look like in the future.