ACT-SQL: In-Context Learning for Text-to-SQL with Automatically-Generated Chain-of-Thought

Hanchong Zhang, Ruisheng Cao, Lu Chen, Hongshen Xu, Kai Yu

Submission Type: Regular Long Paper

Submission Track: Semantics: Lexical, Sentence level, Document Level, Textual Inference, etc.

Submission Track 2: Syntax, Parsing and their Applications

Keywords: text-to-SQL, large language models, in-context learning, chain of thought

Abstract: Recently Large Language Models (LLMs) have been proven to have strong abilities in various domains and tasks. We study the problem of prompt designing in the text-to-SQL task and attempt to improve the LLMs' reasoning ability when generating SQL queries. Besides the trivial few-shot in-context learning setting, we design our chain-of-thought (CoT) prompt with a similar method to schema linking. We provide a method named ACT-SQL to automatically generate auto-CoT exemplars and thus the whole process doesn't need manual labeling. Our approach is cost-saving since we only use the LLMs' API call once when generating one SQL query. Furthermore, we extend our in-context learning method to the multi-turn text-to-SQL task. The experiment results show that the LLMs' performance can benefit from our ACT-SQL approach. Our approach achieves SOTA performance on the Spider dev set among existing in-context learning approaches.

Submission Number: 2903