Teaching Requirements Engineering for AI: A Goal-Oriented Approach in Software Engineering Courses
Project Overview
The document explores the application of generative AI in education, particularly through the lens of Goal-Oriented Requirements Engineering (GORE) for teaching Requirements Engineering in AI-based systems. It emphasizes GORE's effectiveness in capturing high-level requirements essential for AI, as demonstrated in a study involving undergraduate software engineering students. The findings indicate that while students generally appreciate GORE and it effectively teaches fundamental requirements elicitation, there are notable challenges, such as the complexity of managing diagrams and establishing criteria for goal refinement. The document calls for enhancements to the GORE approach to better address these issues and improve educational outcomes, suggesting that with careful refinement, GORE can significantly contribute to the understanding and implementation of AI systems in educational contexts. Overall, the integration of generative AI tools in teaching methodologies shows promise for enriching the learning experience and equipping future engineers with vital skills for developing AI technologies.
Key Applications
Goal-Oriented Requirements Engineering (GORE) using the KAOS method
Context: An introductory software engineering class with 34 undergraduate students
Implementation: Empirical study involving structured lectures, practical exercises, and surveys to evaluate the application of GORE.
Outcomes: 88% of students were able to apply the KAOS method correctly or with minor inadequacies, indicating usability and effectiveness in educational settings.
Challenges: Challenges included determining goal refinement stopping criteria and managing diagram complexity.
Implementation Barriers
Implementation Challenges
Difficulties in determining how to refine goals and knowing when to stop the refinement process.
Proposed Solutions: The study suggests developing clearer stopping criteria and additional support for goal refinement.
Complexity Management
The complexity of assembling and managing GORE diagrams can lead to clutter and confusion.
Proposed Solutions: Additional tools or guidelines to manage diagram complexity are needed.
Prior Knowledge Requirement
Successful application of GORE often requires significant prior knowledge and experience.
Proposed Solutions: More foundational training and resources could help address this barrier.
Project Team
Beatriz Batista
Researcher
Márcia Lima
Researcher
Tayana Conte
Researcher
Contact Information
For information about the paper, please contact the authors.
Authors: Beatriz Batista, Márcia Lima, Tayana Conte
Source Publication: View Original PaperLink opens in a new window
Project Contact: Dr. Jianhua Yang
LLM Model Version: gpt-4o-mini-2024-07-18
Analysis Provider: Openai