Keywords
constraint-aware reranking
schema drift
paired simulation
reproducibility
Abstract
We evaluated constraint-aware reranking for text-to-SQL execution under 21% schema drift. A deterministic paired simulation generated 64 cases and preserved a median slice. Mean execution accuracy changed from 0.484 to 0.537; the paired difference was +0.053 (95% interval +0.050 to +0.055). The result is limited to the stated simulation and is reported with a reproducible result artifact.
References
Jiang, J., Xie, H., Shen, S., Shen, Y., Zhang, Z., Lei, M., Zheng, Y., Li, Y., Li, C., Huang, D., Wu, Y., Zhang, W., Cui, B., & Chen, P. (2025). SiriusBI: A Comprehensive LLM-Powered Solution for Data Analytics in Business Intelligence. Proceedings of the VLDB Endowment, 18(12), 4860-4873. https://doi.org/10.14778/3750601.3750610
C. Joel, U., & Dewole A. Temilola, J. (2025). AN ENHANCED NATURAL LANGUAGE PROCESSING MODEL FOR DATA EXTRACTION AND VISUALIZATION FROM SQL DATABASE STRUCTURES: A SYSTEMATIC REVIEW OF LITERATURE. Jana Nexus: Journal of Computer Science, 01(12), 26-31. https://doi.org/10.21474/jncs01/111
Hairan, B., & Şahman, M. A. (2026). A comparative evaluation of large language models for detecting SQL injection vulnerabilities in web applications. PeerJ Computer Science, 12, e4015. https://doi.org/10.7717/peerj-cs.4015
Petrola, L., Brayner, A., & Franco, W. (2025). Heuristic-Guided Text-to-SQL Translation with LLMs: Optimizing Natural Language Interfaces for Relational Databases. Anais do XL Simpósio Brasileiro de Banco de Dados (SBBD 2025), 126-139. https://doi.org/10.5753/sbbd.2025.247037
Mota, F. D. C., Silva, W. D. V. R. D., & Soares, J. D. N. (2026). BENCHMARK DE MODELOS DE LINGUAGEM DE CÓDIGO ABERTO PARA TEXT-TO-SQL EM DADOS ONCOLÓGICOS BRASILEIROS. REMUNOM, 13(14), 1-67. https://doi.org/10.66104/zvzdrv85
Reichenpfader, D., Müller, H., & Denecke, K. (2023). Large language model-based information extraction from free-text radiology reports: a scoping review protocol. https://doi.org/10.1101/2023.07.28.23292031
Marshan, A., Almutairi, A. N., Ioannou, A., Bell, D., Monaghan, A., & Arzoky, M. (2024). MedT5SQL: a transformers-based large language model for text-to-SQL conversion in the healthcare domain. Frontiers in Big Data, 7, 1371680. https://doi.org/10.3389/fdata.2024.1371680
Ma, X., Tian, X., Wu, L., Wang, X., Tang, X., & Wang, J. (2024). Enhancing Text-to-SQL Capabilities of Large Language Models via Domain Database Knowledge Injection. Frontiers in Artificial Intelligence and Applications. https://doi.org/10.3233/faia240949
Liang, Z., liu, L., Quan, R., zou, M., li, D., Tang, Y., & qin, H. (2026). Hierarchical Adaptive Reward-based Reinforcement Learning Model for High-Precision Text-to-SQL Generation. https://doi.org/10.2139/ssrn.6767042
Öztürk, E. (2025). Improving Text-to-Sql Conversion for Low-Resource Languages Using Large Language Models. Bitlis Eren Üniversitesi Fen Bilimleri Dergisi, 14(1), 163-178. https://doi.org/10.17798/bitlisfen.1561298