OpenSearch-SQL: Enhancing Text-to-SQL with Dynamic Few-shot and Consistency Alignment
Proceedings of the ACM on Management of Data · 17.06.2025 · 3 (3) · 1-24
DOI: 10.1145/3725331 ↗Kurz gefasst
Fachartikel in der Zeitschrift Proceedings of the ACM on Management of Data, online erschienen am 18.06.2025. Verfasst von Xiangjin Xie und 3 weiteren Autor:innen. Die Arbeit wurde bisher 36-mal zitiert (Crossref) und nennt 48 Literaturangaben.
Abstract (Originalsprache)
Although multi-agent collaborative Large Language Models (LLMs) have achieved significant breakthroughs in the Text-to-SQL task, their performance is still constrained by various factors. These factors include the incompleteness of the framework, failure to follow instructions, and model hallucinations. To address these problems, we propose OpenSearch-SQL, which divides the Text-to-SQL task into four main modules: Preprocessing, Extraction, Generation, and Refinement, along with an Alignment module based on a consistency alignment mechanism. This architecture aligns the inputs and outputs of agents through the Alignment module, reducing failures in instruction following and hallucination. Furthermore, we introduce SQL-Like (an intermediate language), optimize the structured Chain-of-Thought (CoT) based on SQL-Like, and develop a dynamic few-shot strategy via self-taught Query-CoT-SQL. In terms of model selection, we directly applied the base LLMs without any post-training, thereby simplifying the task chain and enhancing the framework's portability. Experimental results show that OpenSearch-SQL achieves an execution accuracy(EX) of 69.3% on the BIRD development set, 72.28% on the test set, and a reward-based validity efficiency score (R-VES) of 69.36%, with all three metrics ranking first at the time of submission. These results demonstrate the comprehensive advantages of the proposed method in both effectiveness and efficiency.
Zitieren
APA
Xie, X., Xu, G., Zhao, L., & Guo, R. (2025). OpenSearch-SQL: Enhancing Text-to-SQL with Dynamic Few-shot and Consistency Alignment. Proceedings of the ACM on Management of Data, 3(3), 1-24. https://doi.org/10.1145/3725331
Vancouver
Xie X, Xu G, Zhao L, Guo R. OpenSearch-SQL: Enhancing Text-to-SQL with Dynamic Few-shot and Consistency Alignment. Proceedings of the ACM on Management of Data. 2025;3(3):1-24. doi:10.1145/3725331
BibTeX
@article{Xie2025,
author = {Xie, Xiangjin and Xu, Guangwei and Zhao, Lingyan and Guo, Ruijie},
title = {OpenSearch-SQL: Enhancing Text-to-SQL with Dynamic Few-shot and Consistency Alignment},
journal = {Proceedings of the ACM on Management of Data},
year = {2025},
volume = {3},
number = {3},
pages = {1--24},
doi = {10.1145/3725331}
}
EUVIMED