Bibliographic Details
| Title: |
多维上下文关系感知的SQL 自动生成方法. (Chinese) |
| Alternate Title: |
SQL automatic generation method with multi-dimensional context-relation awareness. (English) |
| Authors: |
刘晨旭, 王邦平, 宋海权, 韩 楠, 杨春芳, 乔少杰 |
| Source: |
Journal of Chongqing University of Technology (Natural Science); 2025, Vol. 39 Issue 9, p124-132, 9p |
| Subject Terms: |
NATURAL language processing, CONTEXT-aware computing, MACHINE learning, APPLIED sciences |
| Abstract (English): |
The generation technique of Structured Query Language (SQL), which automatically converts Natural Language (NL) into SQL, is of keen academic interest. However, current SQL generation methods have several defects: 1) inability of accurately generating SQLs in complex scenarios; 2) inadequate modeling of relationships between NL and database schema elements; and 3) insufficient capability of handling multi-turn dialogues in context-dependent environments. To address these issues, this paper proposes a SQL automatic generation method with multi-dimensional context-relation awareness (MCRA), intergrating such key components as a multi-dimensional relational graph construction module, a multi-dimensional relational-aware encoder, and auxiliary task modules. MCRA enables comprehensive modeling of relationships between elements and partially remedies low accuracy of SQL generation through multi-turn dialogues. Experiments are conducted on benchmark datasets. Results demonstrate MCRA achieves superior SQL generation accuracy compared to other models. [ABSTRACT FROM AUTHOR] |
| Abstract (Chinese): |
结构化查询语言 (structured query language, SQL) 生成技术能将自然语言 (natural language, NL) 自动转换为 SQL, 成为当前研究热点。现有 SQL 自动生成方法仍存在诸多不足: 无法在复杂情况下准确生成 SQL; 无法充分建模 NL 及数据库元素间的关系; 在上下文相关环境下的多轮对话处理能力不足。针对上述问题, 提出多维上下维关系感知的 SQL 自动生成方法 MCRA (multi-dimensional context-relation awareness), 集成了多维关系图构建模块、多维关系感知编码器、辅助任务模块等关键组件, 能够更加全面地建模各元素间的关系, 在一定程度上克服多轮对话生成的 SQL 准确率较低的问题。在标准数据集上进行实验, 结果表明: MCRA 算法 SQL 生成准确率优于主流模型。 [ABSTRACT FROM AUTHOR] |
|
Copyright of Journal of Chongqing University of Technology (Natural Science) is the property of Chongqing University of Technology and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.) |
| Database: |
Complementary Index |
Nájsť tento článok vo Web of Science