大模型增强知识图谱的构建与推理研究综述

doi:10.3778/j.issn.1673-9418.2503034

摘要/Abstract

摘要： 随着知识图谱（KG）在智能问答、推荐系统等领域的广泛应用，其规模化构建与高效推理的技术瓶颈日益凸显。传统人工与半自动化构建模式成本高昂，实体歧义消解、关系抽取准确率不足等问题导致图谱质量难以保障；知识稀疏性与推理规则复杂性限制了图谱推理的泛化能力。大模型（LLM）凭借强大的语义理解与上下文建模能力，为突破这些瓶颈提供了新路径，但目前相关技术缺乏系统性梳理，不同方法的适用场景与性能边界尚不清晰，因此，对现有的大模型增强知识图谱构建和推理的方法开展系统性综述。介绍了知识图谱和大模型相关的基础理论。聚焦知识抽取、自动化构建、知识补全与推理四类核心任务，剖析现有方法：知识抽取中，对比基于大模型的零样本知识抽取与微调领域适配抽取；自动化构建方面，梳理大模型驱动的本体自动生成、图谱迭代更新技术；知识补全方面，总结了借助大模型生成伪三元组、提示规划上下文与结合外部检索机制的方法；推理任务中，分析了基于大模型的静态增强推理和主动规划推理技术。介绍了该研究在医疗、教育等领域的典型应用场景，梳理了支撑相关研究的中英文通用和垂直领域知识图谱数据集资源。针对现存方法的局限性提出了未来研究方向。

关键词: 大模型, 知识图谱, 知识图谱构建, 知识图谱推理

Abstract: With the widespread application of knowledge graphs (KGs) in fields such as intelligent question answering and recommender systems, the technical bottlenecks in large-scale construction and efficient reasoning have become increasingly prominent. Traditional manual or semi-automated construction approaches are costly, while issues such as entity disambiguation and relation extraction accuracy continue to hinder the quality of the resulting graphs. Furthermore, knowledge sparsity and the complexity of reasoning rules limit the generalization capability of KG reasoning. Large language models (LLMs), with their powerful semantic understanding and contextual modeling capabilities, offer promising new avenues to address these challenges. However, current research in this area lacks a systematic review, and the applicability and performance boundaries of various methods remain unclear. To bridge this gap, this paper provides a comprehensive survey of LLM-enhanced knowledge graph construction and reasoning methods. Firstly, this paper introduces the foundational theories of knowledge graphs and large language models. The survey then focuses on four core tasks: knowledge extraction, automated construction, knowledge completion, and reasoning. For knowledge extraction, this paper compares zero-shot extraction methods based on LLMs with domain-adapted extraction through fine-tuning. In terms of automated construction, this paper reviews techniques for LLM-driven ontology generation and iterative graph updates. For knowledge completion, this paper summarizes methods involving pseudo-triple generation via LLMs, prompt-based context planning, and the integration of external retrieval mechanisms. Regarding reasoning tasks, this paper analyzes both static LLM-augmented reasoning and actively planned reasoning approaches. This paper further presents typical application scenarios in domains such as healthcare and education, and compiles a list of general-purpose and domain-specific knowledge graph datasets in both English and Chinese that support research in this area. Finally, this paper highlights the current limitations of existing methods and proposes several future research directions.

Key words: large language models, knowledge graphs, knowledge graph construction, knowledge graph reasoning

张静, 黄文锋, 吴春江, 谭浩. 大模型增强知识图谱的构建与推理研究综述[J]. 计算机科学与探索, 2025, 19(11): 2855-2872.

ZHANG Jing, HUANG Wenfeng, WU Chunjiang, TAN Hao. Overview of Knowledge Graph Construction and Reasoning Enhanced by Large Language Models[J]. Journal of Frontiers of Computer Science and Technology, 2025, 19(11): 2855-2872.

参考文献

[1] ZHAO W X, ZHOU K, LI J Y, et al. A survey of large language models[EB/OL]. [2025-01-13]. https://arxiv.org/abs/2303.18223.
[2] PAN S R, LUO L H, WANG Y F, et al. Unifying large language models and knowledge graphs: a roadmap[EB/OL]. [2025-01-13]. https://arxiv.org/abs/2306.08302.
[3] PAN J Z, RAZNIEWSKI S, KALO J C, et al. Large language models and knowledge graphs: opportunities and challenges[EB/OL]. [2025-01-13]. https://arxiv.org/abs/2308.06374.
[4] LI H, APPLEBY G, SUH A. A preliminary roadmap for LLMs as assistants in exploring, analyzing, and visualizing knowledge graphs[EB/OL]. [2025-01-13]. https://arxiv.org/abs/2404.01425.
[5] IBRAHIM N, ABOULELA S, IBRAHIM A, et al. A survey on augmenting knowledge graphs (KGs) with large language models (LLMs): models, evaluation metrics, benchmarks, and challenges[J]. Discover Artificial Intelligence, 2024, 4(1): 76.
[6] SUN J, XU C, TANG L, et al. Think-on-Graph: deep and responsible reasoning of large language model on knowledge graph[C]//Proceedings of the 11th International Conference on Learning Representations, 2023.
[7] 黄勃, 吴申奥, 王文广, 等. 图模互补: 知识图谱与大模型融合综述[J]. 武汉大学学报(理学版), 2024, 70(4): 397-412.
HUANG B, WU S A, WANG W G, et al. KG-LLM-MCom: a survey on integration of knowledge graph and large language model[J]. Journal of Wuhan University (Natural Science Edition), 2024, 70(4): 397-412.
[8] KHORASHADIZADEH H, AMARA F Z, EZZABADY M, et al. Research trends for the interplay between large language models and knowledge graphs[EB/OL]. [2025-01-13]. https://arxiv.org/abs/2406.08223.
[9] 孔文强. 基于LLM Prompt技术的领域知识图谱构建框架研究[D]. 成都: 电子科技大学, 2024.
KONG W Q. A framework of constructing domain knowledge graph based on LLM prompt[D]. Chengdu: University of Electronic Science and Technology of China, 2024.
[10] VASWANI A. Attention is all you need[C]//Advances in Neural Information Processing Systems 30, 2017: 5998-6008.
[11] SHANAHAN M. Talking about large language models[J]. Communications of the ACM, 2024, 67(2): 68-79.
[12] ACHIAM O J, ADLER S, AGARWAL S, et al. GPT-4 technical report[R]. San Francisco: OpenAI, 2023.
[13] THOPPILAN R, DE FREITAS D, HALL J, et al. LaMDA: language models for dialog applications[EB/OL]. [2025-01-13]. https://arxiv.org/abs/2201.08239.
[14] CHOWDHERY A, NARANG S, DEVLIN J, et al. PaLM: scaling language modeling with pathways[EB/OL]. [2025-01-13]. https://arxiv.org/abs/2204.02311.
[15] YE H B, ZHANG N Y, CHEN H, et al. Generative knowledge graph construction: a review[EB/OL]. [2025-01-13]. https://arxiv.org/abs/2210.12714.
[16] ALAM M, VAN HARMELEN F, ACOSTA M. Towards semantically enriched embeddings for knowledge graph com-pletion[EB/OL]. [2025-01-13]. https://arxiv.org/abs/2308.00081.
[17] GALARRAGA L A, TEFLIOUDI C, HOSE K, et al. AMIE: association rule mining under incomplete evidence in ontological knowledge bases[C]//Proceedings of the 22nd International Conference on World Wide Web. New York: ACM, 2013: 413-422.
[18] YE J J, CHEN X T, XU N, et al. A comprehensive capability analysis of GPT-3 and GPT-3.5 series models[EB/OL]. [2025-01-13]. https://arxiv.org/abs/2303.10420.
[19] WEI X, CUI X Y, CHENG N, et al. Zero-shot information extraction via chatting with ChatGPT[EB/OL]. [2025-01-13]. https://arxiv.org/abs/2302.10205.
[20] POLAK M P, MORGAN D. Extracting accurate materials data from research papers with conversational language models and prompt engineering[J]. Nature Communications, 2024, 15: 1569.
[21] LOU J, LU Y J, DAI D, et al. Universal information extraction as unified semantic matching[J]. Proceedings of the AAAI Conference on Artificial Intelligence, 2023, 37(11): 13318-13326.
[22] LU Y J, LIU Q, DAI D, et al. Unified structure generation for universal information extraction[EB/OL]. [2025-01-14]. https://arxiv.org/abs/2203.12277.
[23] WANG X, ZHOU W K, ZU C, et al. InstructUIE: multi-task instruction tuning for unified information extraction[EB/OL]. [2025-01-14]. https://arxiv.org/abs/2304.08085.
[24] XIAO X L, WANG Y J, XU N, et al. YAYI-UIE: a chat-enhanced instruction tuning framework for universal information extraction[EB/OL]. [2025-01-14]. https://arxiv.org/abs/2312.15548.
[25] ZHONG L F, WU J, LI Q, et al. A comprehensive survey on automatic knowledge graph construction[J]. ACM Computing Surveys, 2024, 56(4): 1-62.
[26] KOMMINENI V K, K?NIG-RIES B, SAMUEL S. From human experts to machines: an LLM supported approach to ontology and knowledge graph construction[EB/OL]. [2025-01-14]. https://arxiv.org/abs/2403.08345.
[27] ZHANG B W, SOH H. Extract, define, canonicalize: an LLM-based framework for knowledge graph construction[EB/OL]. [2025-01-14]. https://arxiv.org/abs/2404.03868.
[28] CHEN B H, BERTOZZI A L. AutoKG: efficient automated knowledge graph generation for language models[C]//Proceedings of the 2023 IEEE International Conference on Big Data. Piscataway: IEEE, 2023: 3117-3126.
[29] CHEN H Z, SHEN X, LV Q T, et al. SAC-KG: exploiting large language models as skilled automatic constructors for domain knowledge graphs[EB/OL]. [2025-01-14]. https://arxiv.org/abs/2410.02811.
[30] CHEN Z, GUO L B, FANG Y, et al. Rethinking uncertainly missing and ambiguous visual modality in multi-modal entity alignment[C]//Proceedings of the 22nd International Semantic Web Conference. Cham: Springer, 2023: 121-139.
[31] ZHANG B, MENDEZ S J R, OMRAN P G. ASKG: an approach to enrich scholarly knowledge graphs through paper decomposition with deep learning[C]//Proceedings of the ISWC 2023 Posters, Demos and Industry Tracks: From Novel Ideas to Industrial Practice Co-Located with the 22nd International Semantic Web Conference, 2023.
[32] YAO L, PENG J Z, MAO C S, et al. Exploring large language models for knowledge graph completion[EB/OL]. [2025-01-14]. https://arxiv.org/abs/2308.13916.
[33] WEI Y B, HUANG Q S, KWOK J T, et al. KICGPT: large language model with knowledge in context for knowledge graph completion[EB/OL]. [2025-01-14]. https://arxiv.org/abs/2402.02389.
[34] LI D W, TAN Z, CHEN T L, et al. Contextualization distillation from large language model for knowledge graph completion[EB/OL]. [2025-01-14]. https://arxiv.org/abs/2402.01729.
[35] ZHANG Y C, CHEN Z, GUO L B, et al. Making large language models perform better in knowledge graph completion[EB/OL]. [2025-01-14]. https://arxiv.org/abs/2310.06671.
[36] LI M Z, YANG C H, XU C J, et al. Retrieval, reasoning, re-ranking: a context-enriched framework for knowledge graph completion[EB/OL]. [2025-01-14]. https://arxiv.org/abs/2411.08165.
[37] CHOUDHARY N, REDDY C K. Complex logical reasoning over knowledge graphs using large language models[EB/OL]. [2025-01-14]. https://arxiv.org/abs/2305.01157.
[38] KIM J, KWON Y, JO Y, et al. KG-GPT: a general framework for reasoning on knowledge graphs using large language models[EB/OL]. [2025-01-15]. https://arxiv.org/abs/2310.11220.
[39] SUN L, WANG X C, LI Y D. Pyramid-driven alignment: pyramid principle guided integration of large language models and knowledge graphs[EB/OL]. [2025-01-15]. https://arxiv.org/abs/2410.12298.
[40] MA J, GAO Z T, CHAI Q, et al. Debate on graph: a flexible and reliable reasoning framework for large language models[EB/OL]. [2025-01-15]. https://arxiv.org/abs/2409.03155.
[41] LUO L H, ZHAO Z C, HAFFARI G, et al. Graph-constrained reasoning: faithful reasoning on knowledge graphs with large language models[EB/OL]. [2025-01-15]. https://arxiv.org/abs/2410.13080.
[42] CUI Y N, SUN Z Q, HU W. A prompt-based knowledge graph foundation model for universal in-context reasoning[EB/OL]. [2025-01-15]. https://arxiv.org/abs/2410.12288.
[43] JIANG J H, ZHOU K, ZHAO W X, et al. KG-agent: an efficient autonomous agent framework for complex reasoning over knowledge graph[EB/OL]. [2025-01-15]. https://arxiv.org/abs/2402.11163.
[44] GHAFAROLLAHI A, BUEHLER M J. SciAgents: automating scientific discovery through multi-agent intelligent graph reasoning[EB/OL]. [2025-01-15]. https://arxiv.org/abs/2409.05556.
[45] ZHAO X J, BLUM M, YANG R, et al. AGENTiGraph: an interactive knowledge graph platform for LLM-based chatbots utilizing private data[EB/OL]. [2025-01-15]. https://arxiv.org/abs/2410.11531.
[46] GAO J Q, ZOU X, AI Y, et al. Graph counselor: adaptive graph exploration via multi-agent synergy to enhance LLM reasoning[EB/OL]. [2025-06-07]. https://arxiv.org/abs/2506.03939.
[47] ALAM F, GIGLOU H B, MALIK K M. Automated clinical knowledge graph generation framework for evidence based medicine[J]. Expert Systems with Applications, 2023, 233: 120964.
[48] JIANG P C, XIAO C, CROSS A, et al. GraphCare: enhancing healthcare predictions with personalized knowledge graphs[EB/OL]. [2025-01-16]. https://arxiv.org/abs/2305.12788.
[49] YU H, ZHOU J, LI L, et al. AIPatient: simulating patients with EHRs and LLM powered agentic workflow[EB/OL]. [2025-01-16]. https://arxiv.org/abs/2409.18924.
[50] ZUO K W, JIANG Y R, MO F, et al. KG4Diagnosis: a hierarchical multi-agent LLM framework with knowledge graph enhancement for medical diagnosis[EB/OL]. [2025-01-16]. https://arxiv.org/abs/2412.16833.
[51] 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.
[52] CHEN S, WEN C, JIN L. Design and implementation of medical intelligent questions answering system based on knowledge graph[C]//Proceedings of the 7th IEEE Information Technology, Networking, Electronic and Automation Control Conference. Piscataway: IEEE, 2024: 1814-1818.
[53] ZHOU G X, HAIHONG E, KUANG Z M, et al. Clinical decision support system for hypertension medication based on knowledge graph[J]. Computer Methods and Programs in Biomedicine, 2022, 227: 107220.
[54] 刘佳, 孙新, 张宇晴. 知识图谱与大语言模型协同的教育资源内容审查[J]. 华东师范大学学报(自然科学版), 2024(5): 57-69.
LIU J, SUN X, ZHANG Y Q. Educational resource content review method based on knowledge graph and large language model collaboration[J]. Journal of East China Normal University (Natural Science), 2024(5): 57-69.
[55] 徐喆, 李振江. 智慧教育视域下高校教学资源与社区教育融合模式的探索[J]. 天津电大学报, 2024, 28(2): 49-55.
XU Z, LI Z J. Exploration of the integration model of university teaching resources and community education under the perspective of smart education[J]. Journal of Tianjin Radio and Television University, 2024, 28(2): 49-55.
[56] 吴哲, 陈希鹏. 融合大语言模型的红色资源价值挖掘与教育应用[J]. 情报科学, 2024, 42(11): 175-182.
WU Z, CHEN X P. Leveraging large language models for value mining and educational application of red resources[J]. Information Science, 2024, 42(11): 175-182.
[57] 张龙, 张翔, 来炳恒, 等. 大模型驱动AI“千问千答”精准服务师生个性化需求[J]. 通信学报, 2024, 45(S2): 277-284.
ZHANG L, ZHANG X, LAI B H, et al. AI thousand questions and answers based on large model accurately serve the personalized needs of teachers and students[J]. Journal on Communications, 2024, 45(S2): 277-284.
[58] 丁克玉. 智能辅助初等数学在线学习平台关键技术研究与实现[D]. 合肥: 中国科学技术大学, 2024.
DING K Y. Research and implementation of key technologies for intelligent-assisted elementary mathematics online learning platforms[D]. Hefei: University of Science and Technology of China, 2024.
[59] DA SILVEIRA M, DELADIENNEE L, ACEM K, et al. Combining knowledge graphs and LLMs for hazardous chemical information management and reuse[C]//Proceedings of the 2024 IEEE International Conference on Bioinformatics and Biomedicine. Piscataway: IEEE, 2024: 6766-6773.
[60] CUI J X, NING M N, LI Z J, et al. Chatlaw: a multi-agent collaborative legal assistant with knowledge graph enhanced mixture-of-experts large language model[EB/OL]. [2025-01-16]. https://arxiv.org/abs/2306.16092.
[61] YU C L, LIU X, MAIA J, et al. COSMO: a large-scale E-commerce common sense knowledge generation and serving system at Amazon[C]//Proceedings of the Companion of the 2024 International Conference on Management of Data. New York: ACM, 2024: 148-160.
[62] 田小贵, 许建新, 张振明. 融合知识图谱与大模型的零件工艺设计知识问答[J]. 计算机技术与发展, 2025, 35(8): 101-109.
TIAN X G, XU J X, ZHANG Z M. Knowledge question-answering method for part process design by integrating knowledge graphs and large language models[J]. Computer Technology and Development, 2025, 35(8): 101-109.
[63] 何培龙, 蒋倜军, 熊少辉, 等. 基于知识增强型大语言模型的路面病害诊断方法[J/OL]. 吉林大学学报(工学版) [2025-03-11]. https://link.cnki.net/doi/10.13229/j.cnki.jdxbgxb.20241146.
HE P L, JIANG T J, XIONG S H, et al. Pavement distress diagnosis based on knowledge-enhanced large language models[J/OL]. Journal of Jilin University (Engineering and Technology Edition) [2025-03-11]. https://link.cnki.net/doi/10.13229/j.cnki.jdxbgxb.20241146.
[64] 刘卓娴, 石拓, 胡啸峰. 基于语言模型和图算法的电信网络诈骗警情要素抽取与模式挖掘[J/OL]. 情报杂志 [2025-03-11]. http://kns.cnki.net/kcms/detail/61.1167.g3.20250304. 1743.004.html.
LIU Z X, SHI T, HU X F. Extraction of elements and pattern mining of telecommunication network fraud alerts based on language model and graph algorithm[J/OL]. Journal of Intelligence [2025-03-11]. http://kns.cnki.net/kcms/detail/61.1167.g3.20250304.1743.004.html.
[65] 魏艳飞. 基于知识图谱模型的汽车故障智能诊断方法[J]. 专用汽车, 2025(2): 87-89.
WEI Y F. Intelligent diagnosis method of automobile faults based on knowledge graph model[J]. Special Purpose Vehicle, 2025(2): 87-89.
[66] 许骏, 潘欣, 佘向飞. 基于知识图谱与大语言模型的文旅知识智能整合方法研究[J]. 吉林师范大学学报(自然科学版), 2025, 46(1): 111-116.
XU J, PAN X, SHE X F. Intelligent integration method of cultural and tourism knowledge based on knowledge graphs and large language models[J]. Journal of Jilin Normal University (Natural Science Edition), 2025, 46(1): 111-116.
[67] AUER S, BIZER C, KOBILAROV G, et al. DBpedia: a nucleus for a web of open data[C]//Proceedings of the 6th International Semantic Web Conference, the 2nd Asian Semantic Web Conference. Berlin: Springer, 2007: 722-735.
[68] SUCHANEK F M, KASNECI G, WEIKUM G. YAGO: a core of semantic knowledge[C]//Proceedings of the 16th International Conference on World Wide Web. New York: ACM, 2007: 697-706.
[69] BOLLACKER K, COOK R, TUFTS P. Freebase: a shared database of structured general human knowledge[EB/OL]. [2025-01-16]. https://cdn.aaai.org/AAAI/2007/AAAI07-355.pdf.
[70] NAVIGLI R, PONZETTO S P. BabelNet: building a very large multilingual semantic network[C]//Proceedings of the Annual Meeting of the Association for Computational Linguistics. Stroudsburg: ACL, 2010: 216-225.
[71] SPEER R, CHIN J, HAVASI C. ConceptNet 5.5: an open multilingual graph of general knowledge[J]. Proceedings of the AAAI Conference on Artificial Intelligence, 2017, 31(1): 4444-4451.