Construction of a Knowledge Graph for Rare Metal Deposits in South China Based on Large Language Models
-
摘要:
以大语言模型为代表的新一代人工智能技术为地学知识的结构化表达与智能推理提供了新机遇.针对地质领域知识体系复杂，非结构化文本语义分散、难以再利用和可视化的问题，本文以华南稀有金属矿床为研究对象，提出了一种融合矿床成因与找矿标志的统一知识图谱构建策略.研究基于DeepSeek R1-32B大语言模型与提示词工程，从大量地质文献中自动抽取并构建了涵盖Li、Be、Nb、Ta等关键稀有矿种的知识图谱.知识图谱及其拓展性分析的结果表明，华南稀有金属成矿与印支期、燕山期岩浆活动密切相关，具有显著的高分异与岩浆热液作用特征；稀有金属元素呈现Li-Be-Nb-Ta-W-Sn的组合异常.综上所述，基于大语言模型构建的知识图谱揭示了华南稀有金属成矿的多阶段成矿机制，阐明了稀有金属矿床在地球化学异常、构造控制及蚀变分带方面的内在联系，为华南及邻区的稀有金属勘查提供了智能化研究方案.
Abstract:The new generation of artificial intelligence technologies, represented by Large Language Models (LLMs), provides new opportunities for the structured representation and intelligent reasoning of geological knowledge. To address the challenges posed by the complexity of geoscientific knowledge systems, as well as the semantic fragmentation, limited reusability, and poor visualizability of unstructured texts, this study proposes a unified strategy for constructing a knowledge graph that integrates deposit genesis and prospecting indicators, taking rare metal deposits in South China as a study object. Based on the DeepSeek R1-32B large language model and prompt engineering, a knowledge graph covering key rare metal elements such as Li, Be, Nb, and Ta, is automatically extracted and constructed. The knowledge graph construction and its extensibility analysis indicate that rare metal mineralization in South China is closely associated with Indosinian and Yanshanian magmatic activities, characterized by significant high-degree fractionation and magmatic-hydrothermal processes. Rare metal elements exhibit a combinatorial anomaly of Li-Be-Nb-Ta-W-Sn. It is concluded that the knowledge graph constructed using LLMs reveals the multi-stage metallogenic mechanisms of rare metals in South China, clarifies the intrinsic relationships among geochemical anomalies, structural controls, and alteration zoning of rare metal deposits, and provides an intelligent research framework for the exploration of rare metals in South China and adjacent regions.
-
Key words:
- knowledge graph /
- large language model /
- rare metal mineralization /
- intelligent exploration /
- South China /
- big data
-
图 3 100个案例在不同模型下的参数对比
Fig. 3. Comparison of model parameters across 100 cases under different models
图 5 华南稀有金属矿床（点）连接关系
Fig. 5. Connectivity diagram of rare metal deposits and occurrences in South China
图 6 华南稀有金属矿床（点）地球化学异常知识图谱
a.地球化学异常元素聚集图；b.元素分异及配分特征图
Fig. 6. Knowledge graph of geochemical anomalies in rare metal deposits and occurrences in South China
图 7 华南稀有金属矿床（点）岩浆及其形成时代知识图谱
Fig. 7. Knowledge graph of magmatism and formation epochs of rare metal deposits and occurrences in South China
图 8 华南稀有金属矿床（点）赋矿岩石知识图谱
Fig. 8. Knowledge graph of host rocks for rare metal deposits and occurrences in South China
图 9 华南稀有金属矿床（点）赋矿岩石知识图谱
Fig. 9. Knowledge graph of host rocks for rare metal deposits and occurrences in South China
图 10 华南稀有金属矿床（点）成矿流体特征知识图谱
Fig. 10. Knowledge graph of ore-forming fluid characteristics in rare metal deposits and occurrences in South China
图 11 和华南稀有金属矿床（点）相关的地质构造知识图谱
a.矿区常见构造类型图；b.断裂构造常见走向与矿体形态图
Fig. 11. Knowledge graph of geological structures related to rare metal deposits and occurrences in South China
图 12 华南稀有金属矿床（点）矿化类型知识图谱
Fig. 12. Knowledge graph of mineralization types in rare metal deposits and occurrences in South China
图 13 华南稀有金属矿床（点）矿物类型知识图谱
Fig. 13. Knowledge graph of mineral types in rare metal deposits and occurrences in South China
表 1 矿床知识本体模型信息表
Table 1. Mineral deposit knowledge ontology model information table
成矿本体因子（0） 地层信息（1） 地层1 地层2 时代信息（2） 区域断裂形成时代 成矿岩体的成岩年龄 岩石时代 岩浆侵入时间 成矿时代 地质构造（3） 区域构造带 深大断裂 次级断裂 深大断裂数量 小断裂数量 深断裂走向 小断裂走向 多数断裂与邻近断裂之间的夹角 背斜 向斜 流体信息（4） 流体类型1 流体类型2 流体富集元素 流体温度 流体性质 矿化类型（5） 矿化类型1 矿化类型2 矿化类型3 含矿岩石类型（6） 含矿岩石1 含矿岩石2 矿物类型（7） 矿物类型1 矿物类型2 矿物组合（8） 矿物组合1 矿物组合2 岩浆信息（9） 岩浆类型 形成岩石 岩浆成因 找矿本体因子（00） 地球物理异常（10） 重力异常 航磁异常 地球化学异常（11） 正异常元素 负异常元素 
下载: 导出CSV
-
Bikel, D. M., Miller, S., Schwartz, R., et al., 1997. March. Nymble: A High-Performance Learning Name-Finder. Fifth Conference on Applied Natural Language Processing, Washington, D. C. . Cheng, Q. M., 2021. What Are Mathematical Geosciences and Its Frontiers? Earth Science Frontiers, 28(3): 6-25 (in Chinese with English abstract). Cracknell, M. J., Reading, A. M., 2014. Geological Mapping Using Remote Sensing Data: A Comparison of Five Machine Learning Algorithms, Their Response to Variations in the Spatial Distribution of Training Data and the Use of Explicit Spatial Information. Computers & Geosciences, 63: 22-33. https://doi.org/10.1016/j.cageo.2013.10.008 Deng, Y. Y., Song, S. C., Fan, J. X., et al., 2024. Paleontology Knowledge Graph for Data-Driven Discovery. Journal of Earth Science, 35(3): 1024-1034. https://doi.org/10.1007/s12583-023-1943-9 Devlin, J., Chang, M. W., Lee, K., et al., 2019. BERT: Pre-Training of Deep Bidirectional Transformers for Language Understanding. NAACL-HLT 2019, Minneapolis. Dong, J., Qiu, Q. J., Xie, Z., et al., 2023. Understanding Table Content for Mineral Exploration Reports Using Deep Learning and Natural Language Processing. Ore Geology Reviews, 156: 105383. https://doi.org/10.1016/j.oregeorev.2023.105383 Grover, A., Leskovec, J., 2016. Node2Vec: Scalable Feature Learning for Networks. The 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. San Francisco. https://doi.org/10.1145/2939672.2939754 Guan, N. N., Song, D. D., Liao, L. J., 2019. Knowledge Graph Embedding with Concepts. Knowledge-Based Systems, 164: 38-44. https://doi.org/10.1016/j.knosys.2018.10.008 Hou, Z. D., Zhao, Z., Liu, Z. J., et al., 2023. Metallogenetic Regularity and Prospecting Direction of Granite Related Li-Be-Nb-Ta Deposits in the Nanling Region, South China. Acta Petrologica Sinica, 39(7): 1950-1972 (in Chinese with English abstract). doi: 10.18654/1000-0569/2023.07.05 Lahitani, A. R., Permanasari, A. E., Setiawan, N. A., 2016. Cosine Similarity to Determine Similarity Measure: Study Case in Online Essay Assessment. 2016 4th International Conference on Cyber and IT Service Management. Bandung. https://doi.org/10.1109/CITSM.2016.7577578 Lawley, C. J. M., Gadd, M. G., Parsa, M., et al., 2023. Applications of Natural Language Processing to Geoscience Text Data and Prospectivity Modeling. Natural Resources Research, 32(4): 1503-1527. https://doi.org/10.1007/s11053-023-10216-1 Li, L. S., Zhou, R. P., Huang, D. G., 2009. Two-Phase Biomedical Named Entity Recognition Using CRFS. Computational Biology and Chemistry, 33(4): 334-338. https://doi.org/10.1016/j.compbiolchem.2009.07.004 Li, X. F., Wei, X. L., Zhu, Y. T., et al., 2021. Rare Metal Deposits in South China: Types, Characteristics, Distribution and Tectonic Setting. Acta Petrologica Sinica, 37(12): 3591-3614 (in Chinese with English abstract). doi: 10.18654/1000-0569/2021.12.02 Lima, L. A., Görnitz, N., Varella, L. E., et al., 2017. Porosity Estimation by Semi-Supervised Learning with Sparsely Available Labeled Samples. Computers & Geosciences, 106: 33-48. https://doi.org/10.1016/j.cageo.2017.05.004 Ma, X. H., Wang, H. H., Lehmann, B., et al., 2024. Control of Magmatic Halogen Composition and Redox State on the Zonation of Metal Mineralization across Active Continental Margins: Perspectives from the World-Class South China Metallogenic Province. Chemical Geology, 669: 122363. https://doi.org/10.1016/j.chemgeo.2024.122363 Ma, X. H., Yan, J. Y., Wang, H. H., et al., 2025. Decoupling of Tungsten and Tin Mineralization Governed by Magmatic Fluorine. Scientia Sinica (Terrae), 55(8): 2583-2602 (in Chinese with English abstract). Ma, X. Z., Hovy, E., 2016. End-to-End Sequence Labeling via Bi-Directional LSTM-CNNS-CRF. arXiv, 1603.01354. https://arxiv.org/abs/1603.01354 Maynez, J., Narayan, S., Bohnet, B., et al., 2020. On Faithfulness and Factuality in Abstractive Summarization. The 58th Annual Meeting of the Association for Computational Linguistics. Online. https://doi.org/10.18653/v1/2020.acl-main.173 Marrero, M., Urbano, J., Sánchez-Cuadrado, S., et al., 2013. Named Entity Recognition: Fallacies, Challenges and Opportunities. Computer Standards & Interfaces, 35(5): 482-489. https://doi.org/10.1016/j.csi.2012.09.004 Ou, Q., Liu, J. Y., Zi, F., et al., 2025. Late Mesozoic Wangxiang Composite Granitic Pluton, South China Block: Implications to Magma Emplacement and Evolution from Geochemical Proxies. Journal of Earth Science, 36(2): 485-507. https://doi.org/10.1007/s12583-022-1760-8 Qin, J. H., Wang, D. H., Wang, Y., et al., 2023. Metallogenci Law and Exploration Prospect of the Middle Part of Nanling Metallogenic Belt, South China Block. Geology in China, Online (in Chinese with English abstract). https://link.cnki.net/urlid/11.1167.P.20230811.1649.002 Qiu, Q., Ma, K., Lü, H., et al., 2023. Construction and Application of a Knowledge Graph for Iron Deposits Using Text Mining Analytics and a Deep Learning Algorithm. Mathematical Geosciences, 55(3): 423-456. https://doi.org/10.1007/s11004-023-10050-4 Song, S. W., Mao, J. W., Romer, R. L., et al., 2024. Hosts of Sn in Reduced Deep-Seated W Skarn Systems: A Case Study on the World-Class Scheelite Skarn Deposit, Zhuxi, South China. Mineralium Deposita, 59(7): 1433-1454. https://doi.org/10.1007/s00126-024-01271-6 Wang, C. B., Ma, X. G., Chen, J. G., et al., 2018. Information Extraction and Knowledge Graph Construction from Geoscience Literature. Computers & Geosciences, 112: 112-120. https://doi.org/10.1016/j.cageo.2017.12.007 Wu, F. Y., Guo, C. L., Hu, F. Y., et al., 2023. Petrogenesis of the Highly Fractionated Granites and Their Mineralizations in Nanling Range, South China. Acta Petrologica Sinica, 39(1): 1-36 (in Chinese with English abstract). doi: 10.18654/1000-0569/2023.01.01 Wu, Q. H., Zhou, H. X., Liu, B., et al., 2023. Spatio-Temporal Distribution of Granite-Related Rare Metal Deposits and W-Sn Deposits in South China and Their Genetic Relationship. Bulletin of Geological Science and Technology, 42(1): 78-88 (in Chinese with English abstract). Xiao, F., Chen, Z. J., Chen, J. G., et al., 2016. A Batch Sliding Window Method for Local Singularity Mapping and Its Application for Geochemical Anomaly Identification. Computers & Geosciences, 90: 189-201. https://doi.org/10.1016/j.cageo.2015.11.001 Xiao, W. Z., Liu, C. Y., Tan, K. X., et al., 2025. Zircon U-Pb-Hf and Trace Element Signatures Reveal the Petrogenesis of the Jiuyishan Granitic Complex, South China: Implications for W-Sn and Rare Metal Mineralization. Journal of Earth Science, 36(3): 1069-1089. https://doi.org/10.1007/s12583-023-1842-2 Yan, Q., Xue, L. F., Li, Y. S., et al., 2023. Mineral Prospectivity Mapping Integrated with Geological Map Knowledge Graph and Geochemical Data: A Case Study of Gold Deposits at Raofeng Area, Shaanxi Province. Ore Geology Reviews, 161: 105651. https://doi.org/10.1016/j.oregeorev.2023.105651 Yang, J. H., Wu, J. H., Zhou, M. F., et al., 2025. Mantle Contributions to Global Tungsten Recycling and Mineralization. Communications Earth & Environment, 6: 510. https://doi.org/10.1038/s43247-025-02471-2 Zhang, R. Z., Zuo, R. G., 2025. Comparison of the Geochemical Characteristics of Highly Fractionated Granites in South China and the Himalaya. Acta Petrologica Sinica, 41(4): 1424-1441 (in Chinese with English abstract). doi: 10.18654/1000-0569/2025.04.16 Zhang, Z. J., Cheng, Q. M., Yang, J., et al., 2021. Machine Learning for Mineral Prospectivity: A Case Study of Iron-Polymetallic Mineral Prospectivity in Southwestern Fujian. Earth Science Frontiers, 28(3): 221-235 (in Chinese with English abstract). Zhang, Z. J., Yang, Z. X., Jian, F. Y., et al., 2026. Interpretability-Enhanced Mineral Prospectivity Models: A Synergistic Approach Using Large Language Models, Knowledge Graphs, and Machine Learning. Mathematical Geosciences, 58(1): 195-231. https://doi.org/10.1007/s11004-025-10231-3 Zhao, M. L., Zhang, Z. J., Yang, J., et al., 2025. Knowledge Graph Construction and Knowledge Discovery for Porphyry Copper Deposits. Ore Geology Reviews, 186: 106875. https://doi.org/10.1016/j.oregeorev.2025.106875 Zhao, Y., He, J. W., Zhu, S. C., et al., 2024. Security of Large Language Models: Current Status and Challenges. Computer Science, 51(1): 68-71 (in Chinese with English abstract). Zhao, Z., Chen, Y. C., Wang, D. H., et al., 2022. Transformation of Mesozoic Dynamic Systems and Superposition of Metallogenic Series of W-Sn-Li-Be-Nb-Ta-REE Mineral Deposits in South China. Acta Petrologica Sinica, 38(2): 301-322 (in Chinese with English abstract). doi: 10.18654/1000-0569/2022.02.01 Zhou, Y. Z., Zuo, R. G., Liu, G., et al., 2021. The Great-Leap-Forward Development of Mathematical Geoscience during 2010-2019: Big Data and Artificial Intelligence Algorithm Are Changing Mathematical Geoscience. Bulletin of Mineralogy, Petrology and Geochemistry, 40(3): 556-573, 777 (in Chinese with English abstract). Zhu, Y. S., 2006. Basic Theory of Mineral Resources Assessment-Theory System between Regional Metallogeny to Mineral Exporation. Acta Geologica Sinica, 80(10): 1518-1527 (in Chinese with English abstract). Zhu, Z. S., Zhu, L., 1998. Prediction Theory and Method System of Deposits. Journal of Chengdu University of Technology (Science & Technology Edition), 25(S1): 6-12 (in Chinese with English abstract). 成秋明, 2021. 什么是数学地球科学及其前沿领域? 地学前缘, 28(3): 6-25. 侯占德, 赵正, 柳振江, 等, 2023. 南岭花岗岩区锂铍铌钽成矿规律与找矿方向. 岩石学报, 39(7): 1950-1972. 李晓峰, 韦星林, 朱艺婷, 等, 2021. 华南稀有金属矿床: 类型、特点、时空分布与背景. 岩石学报, 37(12): 3591-3614. 马星华, 闫金禹, 王宏晖, 等, 2025. 岩浆中的氟对钨锡成矿解耦的控制. 中国科学: 地球科学, 55(8): 2583-2602. 秦锦华, 王登红, 王岩, 等, 2023. 华南板块南岭成矿带中段成矿规律与找矿前景. 中国地质, 知网首发. https://link.cnki.net/urlid/11.1167.P.20230811.1649.002 吴福元, 郭春丽, 胡方泱, 等, 2023. 南岭高分异花岗岩成岩与成矿. 岩石学报, 39(1): 1-36. 吴堑虹, 周厚祥, 刘飚, 等, 2023. 华南与花岗岩有关的稀有金属矿床和钨锡矿床的时空分布规律及其成因联系. 地质科技通报, 42(1): 78-88. 张睿泽, 左仁广, 2025. 华南和喜马拉雅高分异花岗岩地球化学特征对比分析. 岩石学报, 41(4): 1424-1441. 张振杰, 成秋明, 杨玠, 等, 2021. 机器学习与成矿预测: 以闽西南铁多金属矿预测为例. 地学前缘, 28(3): 221-235. 赵月, 何锦雯, 朱申辰, 等, 2024. 大语言模型安全现状与挑战. 计算机科学, 51(1): 68-71. 赵正, 陈毓川, 王登红, 等, 2022. 华南中生代动力体制转换与钨锡锂铍铌钽稀土矿床成矿系列的叠加演化. 岩石学报, 38(2): 301-322. 周永章, 左仁广, 刘刚, 等, 2021. 数学地球科学跨越发展的十年: 大数据、人工智能算法正在改变地质学. 矿物岩石地球化学通报, 40(3): 556-573, 777. 朱裕生, 2006. 矿产预测理论: 区域成矿学向矿产勘查延伸的理论体系. 地质学报, 80(10): 1518-1527. 朱章森, 朱磊, 1998. 矿床定量预测理论与方法体系. 成都理工大学学报(自然科学版), 25(S1): 6-12. -
点击查看大图
计量
- 文章访问数: 204
- HTML全文浏览量: 26
- PDF下载量: 52
- 被引次数: 0
