基于大语言模型的知识图谱构建及应用研究

doi:10.3778/j.issn.1673-9418.2406013

摘要/Abstract

摘要： 海量核电分布式控制系统（DCS）运维数据蕴含着丰富的运维经验和专家知识，如何有效地从中抽取DCS报警响应相关信息并形成知识服务，是目前核电DCS快速响应的热点和前沿研究。由于核电DCS多源异构数据缺乏明确的结构和规范，以往的知识抽取主要依赖人工标注和深度学习的方式进行，但需要具备广泛的领域知识和信息处理能力，且受限于繁重的数据标注工作。提出了分步提示策略的大语言模型知识抽取方法，构建了DCS运维知识图谱；并基于大语言模型技术和二次意图识别方法，利用知识图谱开展智能问答等知识服务。通过以某核电厂DCS运维数据为例，重点就知识抽取、图谱构建、智能问答开展实例研究。结果表明，模型的总体精确率、召回率和F1值分别为91.24%、85.85%和88.43%，能够较为全面地获取DCS多源异构运维数据中的关键实体及属性信息，指导开展领域知识问答，有助于运维人员及时响应DCS报警异常，分析总结故障原因及响应策略，为后期的电厂DCS运维的培训和维护提供借鉴和参考。

关键词: 核电分布式控制系统, 知识图谱, 大语言模型, 知识抽取, 智能问答

Abstract: Massive amounts of operational and maintenance (O&M) data from nuclear power distributed control system (DCS) contain rich operational experience and expert knowledge. Effectively extracting DCS alarm response information and forming knowledge service is a current hotspot and frontier research area in rapid DCS response. Due to the lack of clear structure and standards in multi-source heterogeneous data of nuclear power DCS, previous knowledge extraction primarily relied on manual annotation and deep learning methods, which require extensive domain knowledge and information processing capabilities and are constrained by the heavy workload of data annotation. Therefore, this study proposes a knowledge extraction method using large language model (LLM) with a step-by-step prompting strategy, constructing a DCS O&M knowledge graph (KG). Based on large language model technology and secondary intent recognition methods, intelligent question and answer (Q&A) and other knowledge services are developed utilizing the knowledge graph. Using O&M data from a nuclear power plant’s DCS as a case study, the research focuses on knowledge extraction, knowledge graph construction, and intelligent Q&A. The results show that the model achieves an overall precision (P) of 91.24%, recall (R) of 85.85%, and F1-score of 88.43%. The proposed method can comprehensively capture key entities and attribute information from multi-source heterogeneous DCS O&M data, guiding domain knowledge Q&A, assisting O&M personnel in timely responding to DCS alarm anomalies, analyzing fault causes and response strategies, and providing guidance for DCS O&M training and maintenance in power plants.

Key words: nuclear power distributed control system, knowledge graph, large language model, knowledge extraction, intelligent question and answer (Q&A)

张才科, 李小龙, 郑胜, 蔡家骏, 叶小舟, 罗静. 基于大语言模型的知识图谱构建及应用研究[J]. 计算机科学与探索, 2024, 18(10): 2656-2667.

ZHANG Caike, LI Xiaolong, ZHENG Sheng, CAI Jiajun, YE Xiaozhou, LUO Jing. Research on Construction and Application of Knowledge Graph Based on Large Language Model[J]. Journal of Frontiers of Computer Science and Technology, 2024, 18(10): 2656-2667.

参考文献

[1] 徐颖, 杨武, 刘胜智, 等. 基于数字化系统的核电厂主控室手操器改造方案研究与实践[J]. 核动力工程, 2024, 45(2): 187-192.
XU Y, YANG W, LIU S Z, et al. Research and practice on transformation scheme of hand operator in main control room of nuclear power plant based on digital system[J]. Nuclear Power Engineering, 2024, 45(2): 187-192.
[2] WANG F, WU Y, BU Y, et al. Fault diagnosis of DCS SMPSs in nuclear power plants based on machine learning[J]. Arabian Journal for Science and Engineering, 2023: 1-20.
[3] DAI L, LU W. Operator error types in a dcs of a nuclear power plant[C]//Proceedings of the AHFE 2018 International Conference on Human Error, Reliability, Resilience, and Performance, Orlando, Jul 21-25, 2018. Cham: Springer, 2019: 223-230.
[4] PENG H, WANG Y, ZHANG X, et al. Optimization of preventive maintenance of nuclear safety-class DCS based on reliability modeling[J]. Nuclear Engineering and Technology, 2022, 54(10): 3595-3603.
[5] HE W, NIU Y, DAI S, et al. Study on the system method of DCS remote operation and maintenance system of nuclear power plant based on parallel theory[EB/OL]. [2024-03-10]. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4679098.
[6] TANG C, YU C, GAO Y, et al. Deep learning in nuclear industry: a survey[J]. Big Data Mining and Analytics, 2022, 5(2): 140-160.
[7] 李源, 马新宇, 杨国利, 等. 面向知识图谱和大语言模型的因果关系推断综述[J]. 计算机科学与探索, 2023, 17(10): 2358-2376.
LI Y, MA X Y, YANG G L, et al. Survey of causal inference for knowledge graphs and large language models[J]. Journal of Frontiers of Computer Science and Technology, 2023, 17(10): 2358-2376.
[8] ASHOK D, LIPTON Z C. PromptNER: prompting for named entity recognition[EB/OL]. [2024-03-10]. https://arxiv.org/abs/ 2305.15444.
[9] NGUYEN H L, VU D T, JUNG J J. Knowledge graph fusion for smart systems: a survey[J]. Information Fusion, 2020, 61: 56-70.
[10] LIN J, ZHAO Y, HUANG W, et al. Domain knowledge graph-based research progress of knowledge representation[J]. Neural Computing and Applications, 2021, 33: 681-690.
[11] EHRMANN M, HAMDI A, PONTES E L, et al. Named entity recognition and classification in historical documents: a survey[J]. ACM Computing Surveys, 2023, 56(2): 1-47.
[12] JIANG X, ZHOU W, HOU J. Construction of fault diagnosis system for control rod drive mechanism based on know-ledge graph and bayesian inference[J]. Nuclear Science and Techniques, 2023, 34(2): 21.
[13] SHARMA A, AMRITA, CHAKRABORTY S, et al. Named entity recognition in natural language processing: a systema-tic review[C]//Proceedings of the 2nd Doctoral Symposium on Computational Intelligence. Singapore: Springer, 2022: 817-828.
[14] 荆鑫, 王华峰, 刘潜峰, 等. 基于ELMo-GCN的核电领域命名实体识别[J]. 北京航空航天大学学报, 2022, 48(12): 2556-2565.
JING X, WANG H F, LIU Q F, et al. Named entity recognition in nuclear power field based on ELMo-GCN[J]. Journal of Beijing University of Aeronautics and Astronautics, 2022, 48(12): 2556-2565.
[15] SHI R, WANG Z, LIU Y, et al. Improve on entity recognition method based on BiLSTM-CRF model for the nuclear technology knowledge graph[C]//Proceedings of the 2022 5th International Conference on Pattern Recognition and Artificial Intelligence, Chengdu, Aug 19-21, 2022. Piscataway: IEEE, 2022: 241-246.
[16] MA Z, YAN K, WANG H. BERT-based question answering using knowledge graph embeddings in nuclear power domain[C]//Proceedings of the 2023 26th International Conference on Computer Supported Cooperative Work in Design, Rio de Janeiro, May 24-26, 2023. Piscataway: IEEE, 2023: 267-272.
[17] LIU X, WANG H. Knowledge graph construction and decision support towards transformer fault maintenance[C]//Proceedings of the 2021 IEEE 24th International Conference on Computer Supported Cooperative Work in Design, Dalian, May 5-7, 2021.?Piscataway: IEEE, 2021: 661-666.
[18] WANG Y, CHENG Y, QI Q, et al. IDS-KG: an industrial dataspace-based knowledge graph construction approach for smart maintenance[J]. Journal of Industrial Information Integration, 2024, 38: 100566.
[19] LIU P, GUO Y, WANG F, et al. Chinese named entity recognition: the state of the art[J]. Neurocomputing, 2022, 473: 37-53.
[20] ZHAO W X, ZHOU K, LI J, et al. A survey of large language models[EB/OL]. [2024-03-10]. https://arxiv.org/abs/2303.18223.
[21] 张鹤译, 王鑫, 韩立帆, 等. 大语言模型融合知识图谱的问答系统研究[J]. 计算机科学与探索, 2023, 17(10): 2377-2388.
ZHANG H Y, WANG X, HAN L F, et al. Research on question answering system on joint of knowledge graph and large language models[J]. Journal of Frontiers of Computer Science and Technology, 2023, 17(10): 2377-2388.
[22] 冯钧, 畅阳红, 陆佳民, 等. 基于大语言模型的水工程调度知识图谱的构建与应用[J]. 计算机科学与探索, 2024, 18(6): 1637-1647.
FENG J, CHANG Y H, LU J M, et al. Construction and application of knowledge graph for water engineering scheduling based on large language model[J]. Journal of Frontiers of Computer Science and Technology, 2024, 18(6): 1637-1647.
[23] WANG S, SUN X, LI X, et al. GPT-NER: named entity recognition via large language models[EB/OL]. [2024-03-10].https://arxiv.org/abs/2304.10428.
[24] WEBSON A, PAVLICK E. Do prompt-based models really understand the meaning of their prompts?[EB/OL]. [2024-03-10]. https://arxiv.org/abs/2109.01247.
[25] CHEN J, LU G, PAN Z, et al. Research review of the knowledge graph and its application in power system dispatching and operation[J]. Frontiers in Energy Research, 2022, 10: 896836.
[26] JALILIFARD A, CARIDá V F, MANSANO A F, et al. Semantic sensitive TF-IDF to determine word relevance in documents[C]//Advances in Computing and Network Communications. Singapore: Springer, 2021: 327-337.
[27] MASWADI K, GHANI N A, HAMID S, et al. Human activity classification using decision tree and naive Bayes classifiers[J]. Multimedia Tools and Applications, 2021, 80(14): 21709-21726.
[28] STROBELT H, WEBSON A, SANH V, et al. Interactive and visual prompt engineering for ad-hoc task adaptation with large language models[J]. IEEE Transactions on Visualization and Computer Graphics, 2022, 29(1): 1146-1156.
[29] WANG Y, SHEN S, LIM B Y. RePrompt: automatic prompt editing to refine AI-generative art towards precise expressions[C]//Proceedings of the 2023 CHI Conference on Human Factors in Computing Systems, Hamburg, Apr 23-28, 2023. New York: ACM, 2023: 1-29.
[30] JI S, PAN S, CAMBRIA E, et al. A survey on knowledge graphs: representation, acquisition, and applications[J]. IEEE Transactions on Neural Networks and Learning Systems, 2021, 33(2): 494-514.
[31] REGéCIOVá D, KOLá? D, MILKOVI? M. Pattern matching in YARA: improved Aho-Corasick algorithm[J]. IEEE Access, 2021, 9: 62857-62866.