2026-10-14T23:38:09Z https://www.scidb.cn/oai

10.57760/sciencedb.27850 2025-10-10T14:04:07Z

2025-10-10 BibSQL Dataset doi:10.57760/sciencedb.27850 en The first publicly available Chinese Text-to-SQL dataset specifically designed for the bibliographic domain. The dataset comprises a total of 119 question types, with each type associated with 10 question templates. Half of these templates are manually written, while the other half are automatically generated by LLMs. Bibliographic Search; Conversational Search; Chinese; Text-to-SQL Zhenyu Wang PUBLIC https://creativecommons.org/licenses/by-sa/4.0/ dataset http://www.doi.org/10.1371/journal.pone.0334965 Science Data Bank