Year: 2024
Type: conference-paper
Abstract: SQL is the standard for database query languages and is taught in most introductory database courses. Query languages are illustrated and tested through toy examples: small, accessible, instances of databases. These are not always engaging, but coming up with new examples and questions is time-consuming. Existing research in Computer Science Education has shown that Large Language Models (LLMs) can generate coding exercises. However, this has not been demonstrated for SQL yet but could save teachers much time. In this paper, we study whether it is feasible to have ChatGPT-3.5 generate database schemas and associated SQL questions for teachers through a two-part study.
Authors: Willem Aerts, George Fletcher, Daphne Miedema
Institution: Eindhoven University of Technology
Language: English
Subfield: Computer Science Applications
Field: Computer Science
Domain: Physical Sciences
SDG: Quality education
Open Access status: gold
OpenAlex ID: W4400224630