基于NL2SQL的兵棋数据智能统计分析方法研究

doi:10.16182/j.issn1004731x.joss.22-0559

系统仿真学报 ›› 2023, Vol. 35 ›› Issue (9): 2000-2010.doi: 10.16182/j.issn1004731x.joss.22-0559

基于NL2SQL的兵棋数据智能统计分析方法研究

殷来祥¹^,²(), 李志强¹(), 付琼莹¹

^1.国防大学，北京 100091
^2.中国人民解放军71217部队，山东烟台 265200

收稿日期:2022-04-30 修回日期:2022-08-04 出版日期:2023-09-25 发布日期:2023-09-19
通讯作者: 李志强 E-mail:763426528@qq.com;zhiqiangli@yahoo.com.cn
第一作者简介:殷来祥(1991-)，男，硕士生，研究方向为联合作战体系评估。E-mail：763426528@qq.com

Research on Intelligent Statistical Analysis of Wargaming Data Based on NL2SQL

Yin Laixiang¹^,²(), Li Zhiqiang¹(), Fu Qiongying¹

^1.National Defense University, Beijing 100091, China
^2.PLA 71217 Troops, Yantai 265200, China

Received:2022-04-30 Revised:2022-08-04 Online:2023-09-25 Published:2023-09-19
Contact: Li Zhiqiang E-mail:763426528@qq.com;zhiqiangli@yahoo.com.cn

摘要/Abstract

摘要：

面对海量的兵棋数据，传统界面查询的方式已经无法满足指挥员快速、全面、精准查询数据的要求。通过深入分析兵棋数据特点与主流NL2SQL(natural language to struct query language)模型的缺陷，提出了一套适合兵棋数据智能统计查询的解决方案。针对领域数据集缺乏，提出了一套基于人机协助、动态迭代的兵棋数据集构建方案；针对兵棋查询问句时间敏感的问题，提出了一套“规则+深度学习”的时间表达式识别与规范方法；针对兵棋数据量大提取查询值困难的问题，修改完善了Bridge模型的值提取与SQL生成架构。综合运用以上方案，使兵棋数据查询的精准匹配准确率达到75%以上。

关键词: 兵棋, NL2SQL, 数据集, 时间处理, 统计查询

Abstract:

In the face of massive wargaming data, the traditional interface query method can no longer meet the commander's requirements, i.e., fast, comprehensive, and accurate data querying. Through in-depth analysis of the characteristics of wargaming data and the defects of the mainstream natural language to struct query language (NL2SQL) model, a set of solutions for the intelligent statistical query of wargaming datais presented. Due to the lack of datasets, a wargaming dataset construction scheme based on human-machine assistance and dynamic iteration is provided. In order to solve the time-sensitive problem of wargaming querying, time expression recognition and standardization methods based on "rule + deep learning" are proposed. The value extraction and SQL generation architecture of the Bridge model are modified to facilitate the extraction of query value for a large amount of wargaming data. By comprehensively using the above scheme, the query accuracy of wargaming data is significantly enhanced to more than 75%.

Key words: wargaming, NL2SQL(natural language to struct query language), dataset, time processing, statistical query

中图分类号:

TP31

殷来祥,李志强,付琼莹 . 基于NL2SQL的兵棋数据智能统计分析方法研究[J]. 系统仿真学报, 2023, 35(9): 2000-2010.

Yin Laixiang,Li Zhiqiang,Fu Qiongying . Research on Intelligent Statistical Analysis of Wargaming Data Based on NL2SQL[J]. Journal of System Simulation, 2023, 35(9): 2000-2010.

图/表 8

图1

图2

图3

图4

表1

表2

表3

表4

参考文献 18

1	潘璇, 徐思涵, 蔡祥睿, 等. 基于深度学习的数据库自然语言接口综述[J]. 计算机研究与发展, 2021, 58(9): 1925-1950.
	Pan Xuan, Xu Sihan, Cai Xiangrui, et al. A Survey of Database Natural Language Interface Based on Deep Learning[J]. Computer Research and Development, 2021, 58(9): 1925-1950.
2	Kim H, So B H, Han W S, et al. Natural Language to SQL: Where Are We Today?[J]. Proceedings of the VLDB Endowment, 2020, 13(10): 1737-1750.
3	Zhong V, Xiong Caiming, Socher R. Seq2SQL: Generating Structured Queries from Natural Language Using Reinforcement Learning[EB/OL]. (2017-11-09) [2022-02-05]. .
4	Sun Ningyuan, Yang Xuefeng, Liu Yunfeng. TableQA: a Large-scale Chinese Text-to-SQL Dataset for Table-aware SQL Generation[EB/OL]. (2020-06-10) [2022-01-31]. .
5	Yu Tao, Zhang Rui, Yang Kai, et al. Spider: A Large-scale Human-labeled Dataset for Complex and Cross-domain Semantic Parsing and Text-to-SQL Task[C]//Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing. Stroudsburg, PA, USA: ACL, 2018: 3911-3921.
6	Wang Lijie, Zhang Ao, Wu Kun, et al. DuSQL: A Large-scale and Pragmatic Chinese Text-to-SQL Dataset[C]//Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP). Stroudsburg, PA, USA: ACL, 2020: 6923-6935.
7	Scholak T, Schucher N, Bahdanau D. PICARD: Parsing Incrementally for Constrained Auto-regressive Decoding from Language Models[C]//Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing. Stroudsburg, PA, USA: ACL, 2021: 9895-9901.
8	Guo Jiaqi, Zhan Zecheng, Gao Yan, et al. Towards Complex Text-to-SQL in Cross-domain Database with Intermediate Representation[C]//Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics. Stroudsburg, PA, USA: ACL, 2019: 4524-4535 .
9	邓依依, 邬昌兴, 魏永丰, 等. 基于深度学习的命名实体识别综述[J]. 中文信息学报, 2021, 35(9): 30-45.
	Deng Yiyi, Wu Changxing, Wei Yongfeng, et al. A Review of Named Entity Recognition Based on Deep Learning [J]. Chinese Journal of Information, 2021, 35(9): 30-45.
10	王伟, 赵东岩, 苏婷婷. C-TERN:一种基于CFSA的军事新闻文本时间信息处理算法[J]. 北京大学学报(自然科学版), 2014, 50(1): 9-16.
	Wang Wei, Zhao Dongyan, Su Tingting. C-TERN: A CFSA-based Temporal Information Processing Algorithm for Military News Texts [J]. Journal of Peking University(Natural Science Edition), 2014, 50(1): 9-16.
11	马雷雷, 李宏伟, 魏勇, 等. 基于规则的中文文本时间表达式识别和规范化方法[J]. 信息工程大学学报, 2017, 18(5): 560-565.
	Ma Leilei, Li Hongwei, Wei Yong, et al. A Rule-Based Method for Recognition and Normalization of Chinese Text Temporal Expressions [J]. Journal of the University of Information Engineering, 2017, 18(5): 560-565.
12	Lin X V, Socher R, Xiong Caiming. Bridging Textual and Tabular Data for Cross-Domain Text-to-SQL Semantic Parsing[C]//Findings of the Association for Computational Linguistics: EMNLP 2020. Stroudsburg, PA, USA: ACL, 2020: 4870-4888.
13	Yin Pengcheng, Neubig G, Yih W T, et al. TaBERT: Pretraining for Joint Understanding of Textual and Tabular Data[C]//Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics. Stroudsburg, PA, USA: ACL, 2020: 8413-8426.
14	Johnson J, Douze M, Jégou Hervé. Billion-scale Similarity Search With GPUs[J]. IEEE Transactions on Big Data, 2021, 7(3): 535-547.
15	Mikolov T, Sutskever I, Chen Kai, et al. Distributed Representations of Words and Phrases and Their Compositionality[C]//Proceedings of the 26th International Conference on Neural Information Processing Systems. Red Hook, NY, USA: Curran Associates Inc., 2013: 3111-3119.
16	Vaswani A, Shazeer N, Parmar N, et al. Attention Is All You Need[C]//Proceedings of the 31st International Conference on Neural Information Processing Systems. Red Hook, NY, USA: Curran Associates Inc., 2017: 6000-6010.
17	See A, Liu P J, Manning C D. Get to the Point: Summarization with Pointer-generator Networks[C]// Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics. Stroudsburg, PA, USA: ACL, 2017: 1073-1083.
18	Zhong Ruiqi, Yu Tao, Klein D. Semantic Evaluation for Text-to-SQL with Distilled Test Suites[C]//Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP). Stroudsburg, PA, USA: ACL, 2020: 396-411.

参数	值	参数	值
BERT-wmm	默认参数	损失函数	交叉熵
LSTM隐藏层	400	激活函数	ReLU
注意力头	8	Drop_out	0.3
学习率	5×10^-4	Epoch	10
优化器	Adam	Batch Size	64

模型	DuSQL	BqSQL
Bridge	43.1	15.5
本模型	45.6	18.1

方式	验证集	测试集
BqSQL	66.13	62.9
公开数据集+BqSQL	81.20	75.7

模型	测试集
基础模型	38.3
基础模型+时间规范	51.6
基础模型+值匹配	65.2

[1]	罗俊仁, 张万鹏, 项凤涛, 蒋超远, 陈璟. 智能推演综述：博弈论视角下的战术战役兵棋与战略博弈[J]. 系统仿真学报, 2023, 35(9): 1871-1894.
[2]	陈佳昕, 周国辉, 杨建柏. 基于虚拟仿真方法的手部姿态数据集构建技术[J]. 系统仿真学报, 2023, 35(4): 862-870.
[3]	张大永, 杨镜宇, 吴曦. 兵棋推演空中任务智能预测方法研究[J]. 系统仿真学报, 2023, 35(1): 212-220.
[4]	徐佳乐, 张海东, 赵东海, 倪晚成. 基于卷积神经网络的陆战兵棋战术机动策略学习[J]. 系统仿真学报, 2022, 34(10): 2181-2193.
[5]	杨润青, 吴曦. 面向兵棋系统的卫星导航对抗行动推演模型研究[J]. 系统仿真学报, 2022, 34(10): 2213-2220.
[6]	胡晓峰, 齐大伟. 智能化兵棋系统：下一代需要改变的是什么[J]. 系统仿真学报, 2021, 33(9): 1997-2009.
[7]	司光亚, 王艳正. 新一代大型计算机兵棋系统面临的挑战与思考[J]. 系统仿真学报, 2021, 33(9): 2010-2016.
[8]	唐宇波, 沈弼龙, 师磊, 易星. 下一代兵棋系统模型引擎设计问题研究[J]. 系统仿真学报, 2021, 33(9): 2025-2036.
[9]	贺筱媛, 郭圣明, 吴琳, 李东, 许霄, 李丽. 面向智能化兵棋的认知行为建模方法研究[J]. 系统仿真学报, 2021, 33(9): 2037-2047.
[10]	吴琳, 胡晓峰, 陶九阳, 贺筱媛. 面向智能成长的兵棋推演生态系统[J]. 系统仿真学报, 2021, 33(9): 2048-2058.
[11]	曹占广, 陶帅, 胡晓峰, 何吕龙. 国外兵棋推演及系统研究进展[J]. 系统仿真学报, 2021, 33(9): 2059-2065.
[12]	党建武, 谭凌. 改进果蝇算法优化加权极限学习机的入侵检测[J]. 系统仿真学报, 2021, 33(2): 331-338.
[13]	梁爽, 朱作腾, 贾金原. 基于树型空间的三维树木模型匹配[J]. 系统仿真学报, 2020, 32(4): 582-590.
[14]	汤奋, 张欣, 游雄, 武志强, 李坤伟. 陆军战术兵棋地图设计方法研究[J]. 系统仿真学报, 2019, 31(5): 869-878.
[15]	卢裕秋, 孙金玉, 马世伟. 基于深度卷积神经网络的运动目标检测方法[J]. 系统仿真学报, 2019, 31(11): 2275-2280.

基于NL2SQL的兵棋数据智能统计分析方法研究

Research on Intelligent Statistical Analysis of Wargaming Data Based on NL2SQL

RichHTML

PDF (PC)

可视化

摘要/Abstract

引用本文

使用本文

图/表 8

参考文献 18

相关文章 15

编辑推荐

Metrics

本文评价