@inproceedings{duan-etal-2025-dsqg,
title = "{DSQG}-Syn: Synthesizing High-quality Data for Text-to-{SQL} Parsing by Domain Specific Question Generation",
author = "Duan, Shaoming and
Wu, Youxuan and
Liu, Chuanyi and
Zhang, Yuhao and
Wang, Zirui and
Han, Peiyi and
Yu, Shengyuan and
Yan, Liang and
Liang, Yingwei",
editor = "Chiruzzo, Luis and
Ritter, Alan and
Wang, Lu",
booktitle = "Findings of the Association for Computational Linguistics: NAACL 2025",
month = apr,
year = "2025",
address = "Albuquerque, New Mexico",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2025.findings-naacl.162/",
doi = "10.18653/v1/2025.findings-naacl.162",
pages = "2971--2989",
ISBN = "979-8-89176-195-7",
abstract = "Synthetic data has recently proven effective in enhancing the accuracy of Text-to-SQL parsers. However, existing methods generate SQL queries first by randomly sampling tables and columns based on probability and then synthesize natural language questions (NLQs). This approach often produces a large number of NLQ-SQL pairs that are irrelevant to the target domain and inconsistent in query intent, significantly diminishing the fine-tuning effectiveness of LLMs. In this paper, we introduce DSQG-Syn, a novel text-to-SQL data synthesis framework that based on domain-specific question generation. Specifically, we design a question generation method that creates domain-relevant questions based on predefined question types, ensuring coverage of major SQL operations. Guided by these questions, we synthesize NLQ-SQL pairs that are both domain-relevant and intent-consistent. To further enhance data quality, we filter out noisy samples from the generated pairs. When popular open-source LLMs are fine-tuned on our high-quality synthesized dataset, they achieve significant accuracy improvements, surpassing the performance of closed-source LLM-based approaches. Moreover, we demonstrate that our method outperforms existing state-of-the-art (SOTA) data synthesis techniques."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="duan-etal-2025-dsqg">
<titleInfo>
<title>DSQG-Syn: Synthesizing High-quality Data for Text-to-SQL Parsing by Domain Specific Question Generation</title>
</titleInfo>
<name type="personal">
<namePart type="given">Shaoming</namePart>
<namePart type="family">Duan</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Youxuan</namePart>
<namePart type="family">Wu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Chuanyi</namePart>
<namePart type="family">Liu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yuhao</namePart>
<namePart type="family">Zhang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zirui</namePart>
<namePart type="family">Wang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Peiyi</namePart>
<namePart type="family">Han</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Shengyuan</namePart>
<namePart type="family">Yu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Liang</namePart>
<namePart type="family">Yan</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yingwei</namePart>
<namePart type="family">Liang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2025-04</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Findings of the Association for Computational Linguistics: NAACL 2025</title>
</titleInfo>
<name type="personal">
<namePart type="given">Luis</namePart>
<namePart type="family">Chiruzzo</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Alan</namePart>
<namePart type="family">Ritter</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Lu</namePart>
<namePart type="family">Wang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Albuquerque, New Mexico</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-195-7</identifier>
</relatedItem>
<abstract>Synthetic data has recently proven effective in enhancing the accuracy of Text-to-SQL parsers. However, existing methods generate SQL queries first by randomly sampling tables and columns based on probability and then synthesize natural language questions (NLQs). This approach often produces a large number of NLQ-SQL pairs that are irrelevant to the target domain and inconsistent in query intent, significantly diminishing the fine-tuning effectiveness of LLMs. In this paper, we introduce DSQG-Syn, a novel text-to-SQL data synthesis framework that based on domain-specific question generation. Specifically, we design a question generation method that creates domain-relevant questions based on predefined question types, ensuring coverage of major SQL operations. Guided by these questions, we synthesize NLQ-SQL pairs that are both domain-relevant and intent-consistent. To further enhance data quality, we filter out noisy samples from the generated pairs. When popular open-source LLMs are fine-tuned on our high-quality synthesized dataset, they achieve significant accuracy improvements, surpassing the performance of closed-source LLM-based approaches. Moreover, we demonstrate that our method outperforms existing state-of-the-art (SOTA) data synthesis techniques.</abstract>
<identifier type="citekey">duan-etal-2025-dsqg</identifier>
<identifier type="doi">10.18653/v1/2025.findings-naacl.162</identifier>
<location>
<url>https://aclanthology.org/2025.findings-naacl.162/</url>
</location>
<part>
<date>2025-04</date>
<extent unit="page">
<start>2971</start>
<end>2989</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T DSQG-Syn: Synthesizing High-quality Data for Text-to-SQL Parsing by Domain Specific Question Generation
%A Duan, Shaoming
%A Wu, Youxuan
%A Liu, Chuanyi
%A Zhang, Yuhao
%A Wang, Zirui
%A Han, Peiyi
%A Yu, Shengyuan
%A Yan, Liang
%A Liang, Yingwei
%Y Chiruzzo, Luis
%Y Ritter, Alan
%Y Wang, Lu
%S Findings of the Association for Computational Linguistics: NAACL 2025
%D 2025
%8 April
%I Association for Computational Linguistics
%C Albuquerque, New Mexico
%@ 979-8-89176-195-7
%F duan-etal-2025-dsqg
%X Synthetic data has recently proven effective in enhancing the accuracy of Text-to-SQL parsers. However, existing methods generate SQL queries first by randomly sampling tables and columns based on probability and then synthesize natural language questions (NLQs). This approach often produces a large number of NLQ-SQL pairs that are irrelevant to the target domain and inconsistent in query intent, significantly diminishing the fine-tuning effectiveness of LLMs. In this paper, we introduce DSQG-Syn, a novel text-to-SQL data synthesis framework that based on domain-specific question generation. Specifically, we design a question generation method that creates domain-relevant questions based on predefined question types, ensuring coverage of major SQL operations. Guided by these questions, we synthesize NLQ-SQL pairs that are both domain-relevant and intent-consistent. To further enhance data quality, we filter out noisy samples from the generated pairs. When popular open-source LLMs are fine-tuned on our high-quality synthesized dataset, they achieve significant accuracy improvements, surpassing the performance of closed-source LLM-based approaches. Moreover, we demonstrate that our method outperforms existing state-of-the-art (SOTA) data synthesis techniques.
%R 10.18653/v1/2025.findings-naacl.162
%U https://aclanthology.org/2025.findings-naacl.162/
%U https://doi.org/10.18653/v1/2025.findings-naacl.162
%P 2971-2989
Markdown (Informal)
[DSQG-Syn: Synthesizing High-quality Data for Text-to-SQL Parsing by Domain Specific Question Generation](https://aclanthology.org/2025.findings-naacl.162/) (Duan et al., Findings 2025)
ACL
- Shaoming Duan, Youxuan Wu, Chuanyi Liu, Yuhao Zhang, Zirui Wang, Peiyi Han, Shengyuan Yu, Liang Yan, and Yingwei Liang. 2025. DSQG-Syn: Synthesizing High-quality Data for Text-to-SQL Parsing by Domain Specific Question Generation. In Findings of the Association for Computational Linguistics: NAACL 2025, pages 2971–2989, Albuquerque, New Mexico. Association for Computational Linguistics.