Recent research has developed solutions that allow users to interact with databases using natural language. Due to the inherent complexity of natural language, this is a challenging task. The rise of large language models (LLMs) has provided new opportunities to solve this challenge.

Most research in this area focuses on translating natural language to SQL. A few researchers studied the translation from SQL to natural language, however, no convincing recent evaluation in this area could be found. This research gap encouraged us to construct an application that provides translations and explanations of SQL statements and to investigate the quality of this solution. Using machine and user testing, we demonstrated that students appreciate tools such as this solution, which help them learn and understand SQL and that the quality of both translations and explanations is good overall. Further, we report that prompt engineering (providing additional input such as SQL query and question pairs or DB schemas) does not have a great influence on quality.

Our research contributes to closing the research gap, which is the "absence of a tool that can explain and translate SQL statements into natural language, making database data more accessible". Our solution could be used to support SQL learning and could be developed in a number of directions, as outlined in further work.