Snowflake/SQL-Schema-Retrieval
收藏资源简介:
SQL模式检索基准数据集,用于评估在给定数据库的自然语言问题时检索相关模式元素(如表或列)的任务。该数据集基于五个文本到SQL数据集(Spider、BIRD、BEAVER、LiveSQLBench)构建,支持表级和列级粒度检索,旨在模拟真实场景中从大型数据库模式中定位所需元素的挑战。数据集包含多个子集,每个子集提供模式元素文档、查询和相关判断,采用BEIR/MTEB检索格式,适用于信息检索和自然语言处理研究。
SQL Schema Retrieval Benchmark Dataset: A benchmark dataset for evaluating the task of retrieving relevant schema elements (such as tables or columns) given natural language queries about databases. This dataset is built upon five existing text-to-SQL datasets including Spider, BIRD, BEAVER, and LiveSQLBench, supports retrieval at both table-level and column-level granularity, and aims to simulate the challenges of locating required schema elements from large-scale database schemas in real-world scenarios. The dataset comprises multiple subsets, each containing schema element documents, queries and corresponding relevance judgments, and follows the BEIR/MTEB retrieval format, making it suitable for research in information retrieval and natural language processing.




