Taxonomy-guided method improves fixing errors in natural language to SQL queries
Why, Where, How: Taxonomy-guided Error Grounding for Code Repair in NL2SQL
Artificial Intelligence
Summary
Sometimes, computers translate questions asked in plain English into database queries in SQL, but even if those queries run without errors, they can give the wrong answers. The authors developed a system called TEG that helps fix these wrong queries by understanding what kind of mistake was made, where in the query it occurred, and how to correct it. TEG does this by masking parts of the query and suggesting specific edits, improving the accuracy of fixing SQL queries generated from natural language questions.
What this means in practice
- •For database developers: Use TEG to automatically identify and correct errors in SQL queries created from natural language questions, improving query accuracy and reducing debugging time.
- •For business intelligence teams: Enhance natural language interfaces for data analysis by improving the reliability of automatically generated SQL queries via error-guided correction.
Authors
Suchan Lee, Woomin Song, Hwanjo Yu, Sangwoo Mo
Abstract
SQL queries that large language models write from natural language questions can execute successfully yet produce incorrect results, so execution alone does not reveal what to fix. An error taxonomy says why the query is wrong, but not where to look or how to change it. Existing methods can guide SQL correction through feedback, error reports, or generated plans alongside an unmasked query. We introduce TEG(Taxonomy-guided Error Grounding), which turns a supplied diagnosis into a structured correction input for natural language-to-SQL (NL2SQL) correction. Type-specific rules map each error type to construct classes to reconsider and an edit operation to request. TEG masks the selected constructs in the query when applicable and states that operation in an edit instruction. TEG generates candidate corrections from this input, uses execution feedback to guide candidate selection, and repeats the process one annotation at a time for queries with several errors. On NL2SQL-BUGs, TEG reaches 47.3 single-error execution accuracy and 37.0 overall with Qwen2.5-7B-Instruct. Across the model sizes and thinking modes evaluated in the main comparison, TEG outperforms all evaluated baselines on single-error queries, even when the baselines receive the same error-type annotations. With predicted types, TEG stays above direct LLM correction and ErrorLLM on single-error queries.