Modern Defense Technology ›› 2026, Vol. 54 ›› Issue (3): 190-200.DOI: 10.3969/j.issn.1009-086x.2026.03.017
• PAPERS • Previous Articles
Fengguang ZHOU, Chunyan HU, Yuan ZHOU, Haoyuan ZHANG
Received:2025-08-15
Revised:2025-10-01
Online:2026-06-28
Published:2026-07-03
作者简介:周丰光,男,河南周口人。硕士生,研究方向为指挥信息系统。
CLC Number:
Fengguang ZHOU, Chunyan HU, Yuan ZHOU, Haoyuan ZHANG. Research on Military Equipment Knowledge Extraction Method Based on Retrieval-Augmented Generation[J]. Modern Defense Technology, 2026, 54(3): 190-200.
周丰光, 胡春燕, 周园, 张昊源. 基于检索增强生成的军用装备知识抽取方法研究[J]. 现代防御技术, 2026, 54(3): 190-200.
Add to citation manager EndNote|Ris|BibTeX
URL: https://www.xdfyjs.cn/EN/10.3969/j.issn.1009-086x.2026.03.017
| 设计步骤 | 设计描述 |
|---|---|
| 背景知识 | 本文聚焦于军用装备领域,范围具体到美国航母编队军用装备。在本文的本体模型的构建中,将围绕美航母编队的作战平台、电子装备和作战行动3个角度构建8个类别:美航母编队、航空母舰、护航舰艇、舰载机、武器系统、传感器与电子战、指挥与控制和作战行动。上述列举出的领域中的类别需要进一步把它们进行细分子类,其中护航舰艇的子类有巡洋舰、驱逐舰、护卫舰、核潜艇、补给舰;武器系统的子类… |
| 任务描述 | 假设你是一个知识抽取的专家,请结合背景知识、原文本内容、输出要求和参考示例抽取出三元组数据。 |
| 输入文本 | 星座号航空母舰装载了3座8联装“海麻雀”舰对空导弹发射装置;4座SRBOC电子对抗诱饵发射装置。… |
| 输出要求 | 输出格式必须为json格式:[[“主语”,“关系”,“宾语”],…]。关系类型包括: -属性关系 1)美航母编队的属性定义包括名称、隶属编队、母港位置、部署状态等。…其他类别的关系定义如下:从不同维度进行考量:物理部署、功能依赖、任务协同、技术支撑、后勤保障、信息流动、其他关系等。 a)物理部署关系。物理部署关系是指某一类别的实体(如武器、设备)须搭载在另一类别的实体(如舰船、舰载机)上,具体的定义包括“装备”“搭载”“配备”等。例如F-18战机部署在航空母舰上。 b)功能依赖关系。功能依赖是… |
| 参考示例 | [[“星座号航空母舰”,“搭载”,“海麻雀舰对空导弹发射装置”],[“星座号航空母舰”,“搭载”,“SRBOC电子对抗诱饵发射装置”]] |
Table1 Prompt design for knowledge extraction in the field of military equipment
| 设计步骤 | 设计描述 |
|---|---|
| 背景知识 | 本文聚焦于军用装备领域,范围具体到美国航母编队军用装备。在本文的本体模型的构建中,将围绕美航母编队的作战平台、电子装备和作战行动3个角度构建8个类别:美航母编队、航空母舰、护航舰艇、舰载机、武器系统、传感器与电子战、指挥与控制和作战行动。上述列举出的领域中的类别需要进一步把它们进行细分子类,其中护航舰艇的子类有巡洋舰、驱逐舰、护卫舰、核潜艇、补给舰;武器系统的子类… |
| 任务描述 | 假设你是一个知识抽取的专家,请结合背景知识、原文本内容、输出要求和参考示例抽取出三元组数据。 |
| 输入文本 | 星座号航空母舰装载了3座8联装“海麻雀”舰对空导弹发射装置;4座SRBOC电子对抗诱饵发射装置。… |
| 输出要求 | 输出格式必须为json格式:[[“主语”,“关系”,“宾语”],…]。关系类型包括: -属性关系 1)美航母编队的属性定义包括名称、隶属编队、母港位置、部署状态等。…其他类别的关系定义如下:从不同维度进行考量:物理部署、功能依赖、任务协同、技术支撑、后勤保障、信息流动、其他关系等。 a)物理部署关系。物理部署关系是指某一类别的实体(如武器、设备)须搭载在另一类别的实体(如舰船、舰载机)上,具体的定义包括“装备”“搭载”“配备”等。例如F-18战机部署在航空母舰上。 b)功能依赖关系。功能依赖是… |
| 参考示例 | [[“星座号航空母舰”,“搭载”,“海麻雀舰对空导弹发射装置”],[“星座号航空母舰”,“搭载”,“SRBOC电子对抗诱饵发射装置”]] |
| 实验模型 | 方法1 | 方法2 | 方法3 |
|---|---|---|---|
| 稀疏模型 | 无 | 无 | TF-IDF+BM25 |
| 稠密模型(文本嵌入) | 无 | paraphrase-multilingual-MiniLM-L12-v2 | paraphrase-multilingual-MiniLM-L12-v2 |
| 文本生成模型 | DeepSeek-V3 | DeepSeek-V3 | DeepSeek-V3 |
| 调优参数 | 无 | 无 |
Table2 Experimental method configurations
| 实验模型 | 方法1 | 方法2 | 方法3 |
|---|---|---|---|
| 稀疏模型 | 无 | 无 | TF-IDF+BM25 |
| 稠密模型(文本嵌入) | 无 | paraphrase-multilingual-MiniLM-L12-v2 | paraphrase-multilingual-MiniLM-L12-v2 |
| 文本生成模型 | DeepSeek-V3 | DeepSeek-V3 | DeepSeek-V3 |
| 调优参数 | 无 | 无 |
| 数据 | 方法1 | 方法2 | 方法3 |
|---|---|---|---|
| 总数 | 30 723 | 39 350 | 44 792 |
| 第1组 | 13 070 | 15 571 | 16 022 |
| 第2组 | 12 512 | 17 111 | 20 166 |
| 第3组 | 5 141 | 6 668 | 8 604 |
Table3 Triplet data extraction results under three methods
| 数据 | 方法1 | 方法2 | 方法3 |
|---|---|---|---|
| 总数 | 30 723 | 39 350 | 44 792 |
| 第1组 | 13 070 | 15 571 | 16 022 |
| 第2组 | 12 512 | 17 111 | 20 166 |
| 第3组 | 5 141 | 6 668 | 8 604 |
| 实验编号 | 1 | 2 | 3 |
|---|---|---|---|
| 准确率 | 0.773 2 | 0.812 8 | 0.822 3 |
| 召回率 | 0.588 3 | 0.731 6 | 0.901 8 |
| F1分数 | 0.668 3 | 0.770 1 | 0.860 2 |
| 抽取三元组数量 | 7 012 | 8 296 | 10 108 |
| 正确抽取数量 | 5 422 | 6 743 | 8 312 |
| 理应抽取数量 | 9 217 | 9 217 | 9 217 |
Table4 Comparison of evaluation metrics of evaluation data under three models
| 实验编号 | 1 | 2 | 3 |
|---|---|---|---|
| 准确率 | 0.773 2 | 0.812 8 | 0.822 3 |
| 召回率 | 0.588 3 | 0.731 6 | 0.901 8 |
| F1分数 | 0.668 3 | 0.770 1 | 0.860 2 |
| 抽取三元组数量 | 7 012 | 8 296 | 10 108 |
| 正确抽取数量 | 5 422 | 6 743 | 8 312 |
| 理应抽取数量 | 9 217 | 9 217 | 9 217 |
| [1] | SINGHAL K, AZIZI S, TU Tao, et al. Large Language Models Encode Clinical Knowledge[J]. Nature, 2023, 620(7972): 172-180. |
| [2] | CHALKIDIS I, PASINI T, ZHANG Sheng, et al. FairLex: A Multilingual Benchmark for Evaluating Fairness in Legal Text Processing[C]∥Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics. Stroudsburg: ACL, 2022: 4389-4406. |
| [3] | SARMAH B, MEHTA D, PASQUALI S, et al. Towards Reducing Hallucination in Extracting Information from Financial Reports Using Large Language Models[C]∥Proceedings of the Third International Conference on AI-ML Systems. New York, NY, USA: Association for Computing Machinery, 2023: 39. |
| [4] | 钟逸, 冷彦, 陈思慧, 等. 基于大语言模型RAG架构的电池加速研究: 现状与展望[J]. 储能科学与技术, 2024, 13(9): 3214-3225. |
| ZHONG Yi, LENG Yan, CHEN Sihui, et al. Accelerating Battery Research with Retrieval-Augmented Large Language Models: Present and Future[J]. Energy Storage Science and Technology, 2024, 13(9): 3214-3225. | |
| [5] | JI Ziwei, LEE N, FRIESKE R, et al. Survey of Hallucination in Natural Language Generation[J]. ACM Computing Surveys, 2023, 55(12): 248. |
| [6] | TONMOY S M T I, ZAMAN S M M, JAIN V, et al. A Comprehensive Survey of Hallucination Mitigation Techniques in Large Language Models[EB/OL]. (2024-01-08) [2025-09-16]. . |
| [7] | LEWIS P, PEREZ E, PIKTUS A, et al. Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks[C]∥Proceedings of the 34th International Conference on Neural Information Processing Systems. Red Hook: Curran Associates Inc., 2020: 9459-9474. |
| [8] | LI Yunqing, STARLY B. Building a Knowledge Graph to Enrich ChatGPT Responses in Manufacturing Service Discovery[J]. Journal of Industrial Information Integration, 2024, 40: 100612. |
| [9] | 王柳苏. 检索增强生成在医疗问答领域的应用研究[D]. 上海: 华东师范大学, 2024. |
| WANG Liusu. Exploration of Retrieval Augmentation Generation Application in the Medical Question and Answer Domain[D]. Shanghai: East China Normal University, 2024. | |
| [10] | SELVA BIRUNDA S, KANNIGA DEVI R. A Review on Word Embedding Techniques for Text Classification[C]∥Innovative Data Communication Technologies and Application. Singapore: Springer Singapore, 2021: 267-281. |
| [11] | DO Q, MORIYANI M A, LE C, et al. Cost-Weighted TF-IDF: A Novel Approach for Measuring Highway Project Similarity Based on Pay Items’ Cost Composition and Term Frequency[J]. Journal of Construction Engineering and Management, 2023, 149(8): 04023069. |
| [12] | LEWIS P, PEREZ E, PIKTUS A, et al. Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks[C]∥Proceedings of the 34th International Conference on Neural Information Processing Systems. Red Hook: Curran Associates Inc., 2020: 9459-9474. |
| [13] | GAO Yunfan, XIONG Yun, GAO Xinyu, et al. Retrieval-Augmented Generation for Large Language Models: A Survey[EB/OL]. (2024-03-27) [2025-09-16]. . |
| [14] | CHEN Jiawei, LIN Hongyu, HAN Xianpei, et al. Benchmarking Large Language Models in Retrieval-augmented Generation[J]. Proceedings of the AAAI Conference on Artificial Intelligence, 2024, 38(16): 17754-17762. |
| [15] | NOY N F, MCGUINNESS D L. Ontology Development: A Guide to Creating your First Ontology[EB/OL]. [2025-09-16]. . |
| [16] | SANG E F T K, DE MEULDER F. Introduction to the CoNLL-2003 Shared Task: Language-Independent Named Entity Recognition[C]∥Proceedings of the Seventh Conference on Natural Language Learning at HLT-NAACL 2003. Stroudsburg: ACL, 2003: 142-147. |
| [17] | REIMERS N, GUREVYCH I. Making Monolingual Sentence Embeddings Multilingual Using Knowledge Distillation[C]∥Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP). Stroudsburg: ACL, 2020: 4512-4525. |
| [18] | CHEN Wenhu, VERGA P, DE JONG M, et al. Augmenting Pre-trained Language Models with QA-Memory for Open-Domain Question Answering[C]∥Proceedings of the 17th Conference of the European Chapter of the Association for Computational Linguistics. Stroudsburg: ACL, 2023: 1597-1610. |
| [19] | SHI Peng, LIN J. Cross-Lingual Relevance Transfer for Document Retrieval[EB/OL]. (2019-11-08) [2025-09-16]. . |
| [20] | GAO Luyu, DAI Zhuyun, CALLAN J. COIL: Revisit Exact Lexical Match in Information Retrieval with Contextualized Inverted List[C]∥Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies. Stroudsburg: ACL, 2021: 3030-3042. |
| [21] | JUVEKAR K, PURWAR A. COS-Mix: Cosine Similarity and Distance Fusion for Improved Information Retrieval[EB/OL]. (2024-06-02) [2025-09-16]. . |
| [22] | KADHIM A I. Term Weighting for Feature Extraction on Twitter: A Comparison Between BM25 and TF-IDF[C]∥2019 International Conference on Advanced Science and Engineering (ICOASE). Piscataway: IEEE, 2019: 124-128. |
| [23] | HE Chuanni, HE Weilin, LIU Min, et al. Enriched Construction Regulation Inquiry Responses: A Hybrid Search Approach for Large Language Models[J]. Journal of Management in Engineering, 2025, 41(3): 04025001. |
| [24] | 鞠炜刚, 汪鹏, 王佳. 基于大语言模型和RAG的持续交付智能问答系统[J]. 计算机技术与发展, 2025, 35(2): 107-114. |
| JU Weigang, WANG Peng, WANG Jia. Continuous Delivery Intelligent Question-Answering System Based on Large Language Models and RAG[J]. Computer Technology and Development, 2025, 35(2): 107-114. |
| [1] | Tianyu JING, Zehui LIU, Qin CAO, Yanzhao NI, Huifeng KANG, Guangqing XIA. Simulation on the Composite Aerodynamic Characteristics of Gas Jet and Rudder of a Flying Vehicle [J]. Modern Defense Technology, 2026, 54(3): 131-139. |
| [2] | Li ZHENG, Hanpu YU, Qixin ZHANG, Shuanghou DENG, Guoxu FENG, Yannian YANG. Integrated Design of Aerodynamic Stealth for UCAV Considering Inlet and Exhaust System [J]. Modern Defense Technology, 2026, 54(3): 115-130. |
| [3] | Zhen ZHAO, Xiaolin ZHAO, Maolin GAO, Jie MIN, Weikang CHEN. Overview of Military Simulation Experiment Technology [J]. Modern Defense Technology, 2026, 54(2): 34-47. |
| [4] | Yaoyao LI, Ping DONG, Chaohu XU. Aerodynamic Design of Conventional Small Multi-wing Aircraft [J]. Modern Defense Technology, 2025, 53(6): 60-68. |
| [5] | Jingli HUANG, Yuliang ZHAO, Weijie FAN, Qiang XU. Vector Magnetometer Calibration in Unknown Environment of Strong Noise [J]. Modern Defense Technology, 2025, 53(3): 95-102. |
| [6] | Yukai PAN, Wenhao YU, Zhen TANG, Jiayi YIN, Wen ZHANG, Guoxu FENG, Yujie GUO. Node Importance Evaluation of Dynamic Reconfiguration System Based on Unit Attribute [J]. Modern Defense Technology, 2025, 53(1): 108-119. |
| [7] | Xiaoying HE, Yanzhong Lin. Interior Ballistic Design of a Multi-gas Generators Ejecting System [J]. Modern Defense Technology, 2024, 52(6): 41-46. |
| [8] | Yonghua ZENG, Haiyan WANG, Tingting WU, Xiaolong WANG, Baicun JIANG, Hengyuan LIANG. Evaluation of Contingent-Level Battlefield Repair Capability Based on Cloud Model [J]. Modern Defense Technology, 2024, 52(5): 116-126. |
| [9] | Zhigang HU, Jingjun LOU, Yuedong SHI. Optimal Prescheduling of Equipment Support Based on Two-Layer Programming [J]. Modern Defense Technology, 2024, 52(5): 147-155. |
| [10] | Jisong FAN, Yuyao TONG, Jun GAO, Chengxiang TIAN. Combat Efficiency Evaluation of Army Synthetic Operation Unit Based on DEA-Malmquist Index Model [J]. Modern Defense Technology, 2024, 52(5): 17-24. |
| [11] | Yanjun QI, Yufeng WANG, Rui LI. Integrated Support Design Method of Weapons Based on Health Management [J]. Modern Defense Technology, 2023, 51(5): 104-109. |
| [12] | Yebo REN, Lina YANG, Song YAN. Weapon Launcher Creep During Long-Term Storage [J]. Modern Defense Technology, 2023, 51(5): 110-117. |
| [13] | Yaohua ZHANG, Yaning XU, Shaojie GAO, Peng ZHANG, Renti LIU, Xiaolong SHI, Xi YANG. Multi-ballistic Trajectory Planning and Simulation Based on Multi-objective Optimization Algorithm [J]. Modern Defense Technology, 2023, 51(5): 77-85. |
| [14] | Mingnan TANG, Chenglong ZHANG, Qiang ZHAO, Linlin LI. Scenario-Driven and Intelligent Optimization of Disposition Scheme for Air Defense [J]. Modern Defense Technology, 2023, 51(3): 1-9. |
| [15] | Chenglong ZHANG, Qiang ZHAO, Ran WEI, Tong LI. Research on Design Method of Multi-dimensional Architecture of Cross-Domain Operational Systems [J]. Modern Defense Technology, 2023, 51(3): 20-30. |
| Viewed | ||||||
|
Full text |
|
|||||
|
Abstract |
|
|||||