FollowUp
收藏arXiv2025-09-30 收录
下载链接：
https://github.com/siviltaram/followup
下载链接
链接失效反馈官方服务：
资源简介：
该数据集是为研究和评估后续查询分析而设计的高质量数据集，包含了1000组查询三元组，覆盖了120张表格。该数据集拥有大约2000个词汇量，并包含了多种后续查询场景。为了降低对SQL专业知识的依赖，数据集以自然语言形式收集。规模上，该数据集包含了1000组查询三元组，跨越了120张表格，任务专注于后续查询分析。
This high-quality dataset is designed for research and evaluation of subsequent query analysis. It comprises 1,000 query triplets and covers 120 tables, with a vocabulary size of approximately 2,000 words and a wide range of subsequent query scenarios. To minimize reliance on professional SQL knowledge, the dataset is collected in natural language format. In terms of its scale, this dataset includes 1,000 query triplets across 120 tables, and its core task focuses on subsequent query analysis.
提供机构：
Crowdsourcing (8 workers)
