SemanticAgent: A Semantics-Aware Framework for Text-to-SQL Data Synthesis

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Gao, Qiang, Li, Zhenping, Zhuo, Anqi, Zhao, Yingxiao, Geng, Weibo, Li, Xiaosong
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911618312437760
author Gao, Qiang
Li, Zhenping
Zhuo, Anqi
Zhao, Yingxiao
Geng, Weibo
Li, Xiaosong
author_facet Gao, Qiang
Li, Zhenping
Zhuo, Anqi
Zhao, Yingxiao
Geng, Weibo
Li, Xiaosong
contents Existing text-to-SQL synthesis pipelines still conflate executability with semantic validity: syntactic checks and execution-based validation can retain queries that execute successfully while violating database semantics. To address these limitations, we propose SemanticAgent, a semantic-aware synthesis framework. SemanticAgent organizes synthesis around three specialized modules: an analyzer, a synthesizer, and a verifier. Through a three-stage protocol of semantic analysis, stepwise synthesis, and diagnostic refinement, SemanticAgent transforms execution-based validation alone into a traceable reasoning process. Our framework generates synthetic data that consistently outperforms prior synthesis methods under semantic-quality evaluation, leading to stronger downstream fine-tuning performance, especially on semantically demanding benchmarks.
format Preprint
id arxiv_https___arxiv_org_abs_2604_21414
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle SemanticAgent: A Semantics-Aware Framework for Text-to-SQL Data Synthesis
Gao, Qiang
Li, Zhenping
Zhuo, Anqi
Zhao, Yingxiao
Geng, Weibo
Li, Xiaosong
Artificial Intelligence
I.2.7; I.2.4
Existing text-to-SQL synthesis pipelines still conflate executability with semantic validity: syntactic checks and execution-based validation can retain queries that execute successfully while violating database semantics. To address these limitations, we propose SemanticAgent, a semantic-aware synthesis framework. SemanticAgent organizes synthesis around three specialized modules: an analyzer, a synthesizer, and a verifier. Through a three-stage protocol of semantic analysis, stepwise synthesis, and diagnostic refinement, SemanticAgent transforms execution-based validation alone into a traceable reasoning process. Our framework generates synthetic data that consistently outperforms prior synthesis methods under semantic-quality evaluation, leading to stronger downstream fine-tuning performance, especially on semantically demanding benchmarks.
title SemanticAgent: A Semantics-Aware Framework for Text-to-SQL Data Synthesis
topic Artificial Intelligence
I.2.7; I.2.4
url https://arxiv.org/abs/2604.21414