基于知识图谱的碳酸岩型稀土矿成矿要素挖掘

doi:10.13745/j.esf.sf.2025.3.31

摘要/Abstract

摘要：

稀土元素在现代高科技产业中发挥着关键作用,而碳酸岩型稀土矿作为重要的资源类型,其成矿特征和机制尚不明确,严重制约了找矿勘查的突破。随着地学大数据时代的到来,知识图谱技术在矿产资源预测中的应用逐渐成为热点。本研究结合自然语言处理技术和知识图谱构建方法,针对碳酸岩型稀土矿床的成矿特征和机制开展系统研究。通过收集白云鄂博矿床和冕宁—德昌成矿带的相关文献,利用BERT-BiLSTM-CRF实体识别模型和BERT关系抽取模型构建了碳酸岩型稀土矿床领域的知识图谱。研究结果表明,矿物、岩石和元素是成矿的关键节点,其中萤石表现出较强的一致性,具有显著的指示矿物潜力;铕和铈元素因其氧化还原异常显著关联,是重要的找矿指标。此外,不同区域知识图谱反映了成矿类型的差异,白云鄂博矿床以碳酸岩为主,而冕宁—德昌成矿带与碱性岩关系密切。图谱的层次聚类分析进一步揭示了成矿要素间的关联性,为稀土矿床的成矿环境和成矿机理研究提供了新的视角,同时也为找矿预测和资源评价提供了科学依据。

关键词: 知识图谱, 碳酸岩, 稀土矿, 白云鄂博, 冕宁—德昌

Abstract:

Rare earth elements (REEs) are critical for modern high-tech industries, yet the metallogenic characteristics and mechanisms of carbonatite-hosted REE deposits remain poorly understood, significantly hindering exploration breakthroughs. With the advent of the geoscience big data era, knowledge graph technology has emerged as a key tool for mineral resource prediction. This study integrates natural language processing (NLP) techniques and knowledge graph construction methods to systematically investigate the metallogenic characteristics and mechanisms of carbonatite-hosted REE deposits. We collected literature pertaining to the Bayan Obo deposit and the Mianning-Dechang metallogenic belt. Using the BERT-BiLSTM-CRF model for entity recognition and the BERT model for relationship extraction, we constructed a domain-specific knowledge graph. Results indicate that minerals, rocks, and elements are critical nodes influencing mineralization. Fluorite exhibits high consistency across regional knowledge graphs, highlighting its potential as a prospecting indicator mineral. Europium and cerium, due to their redox-sensitive anomalies, serve as important indicators for REE exploration. Calcite and bastnaesite also demonstrate indicative potential. The knowledge graph reveals that Bayan Obo exhibits stronger associations with carbonatites, while the Mianning-Dechang belt is more closely linked to alkaline rocks. Hierarchical clustering further demonstrates significant correlations among similar nodes, providing key insights into the metallogenic environment and critical elements. This study offers a novel perspective and methodology for understanding metallogenic mechanisms and provides robust scientific support for the exploration and evaluation of REE deposits.

Key words: knowledge graph, carbonatite, rare earth deposits, Bayan Obo, Mianning-Dechang

中图分类号:

冯婷婷, 蔡诗柔, 张振杰. 基于知识图谱的碳酸岩型稀土矿成矿要素挖掘[J]. 地学前缘, 2025, 32(4): 262-279.

FENG Tingting, CAI Shirou, ZHANG Zhenjie. Mining elements of carbonatite-type rare earth deposits based on knowledge map[J]. Earth Science Frontiers, 2025, 32(4): 262-279.

图/表 19

参考文献 78

[1]	ZHU Y Q, TAN Y J, ZHANG J T, et al. A framework of Hadoop based Geology Big Data fusion and mining technologies[J]. Acta Geodaetica et Cartographica Sinica, 2015, 44: 152-159.
[2]	周成虎, 王华, 王成善, 等. 大数据时代的地学知识图谱研究[J]. 中国科学: 地球科学, 2021, 51(7): 1070-1079.
[3]	吴冲龙, 刘刚. 大数据与地质学的未来发展[J]. 地质通报, 2019, 38(7): 1081-1088.
[4]	GUO H D. Big data drives the development of Earth science[J]. Big Earth Data, 2017, 1(1/2): 1-3.
[5]	LI S, CHEN J P, LIU C. Overview on the development of intelligent methods for mineral resource prediction under the background of geological big data[J]. Minerals, 2022, 12(5): 616.
[6]	周永章, 黎培兴, 王树功, 等. 矿床大数据及智能矿床模型研究背景与进展[J]. 矿物岩石地球化学通报, 2017, 36(2): 327-331, 344.
[7]	诸云强, 孙凯, 李威蓉, 等. 地球科学知识图谱比较分析与启示: 构建方法与内容视角[J]. 高校地质学报, 2023, 29(3): 382-394.
[8]	ZHOU C H, WANG H, WANG C S, et al. Geoscience knowledge graph in the big data era[J]. Science China: Earth Sciences, 2021, 64(7): 1105-1114.
[9]	齐浩, 董少春, 张丽丽, 等. 地球科学知识图谱的构建与展望[J]. 高校地质学报, 2020, 26(1): 2-10. DOI
[10]	PEI Y, CHAI S L, LI X L, et al. Construction and application of a knowledge graph for gold deposits in the Jiapigou gold metallogenic belt, Jilin Province, China[J]. Minerals, 2022, 12(9): 1173.
[11]	ZHANG X Y, HUANG Y, ZHANG C J, et al. Geoscience knowledge graph (GeoKG): development, construction and challenges[J]. Transactions in GIS, 2022, 26(6): 2480-2494.
[12]	邱芹军, 田苗, 吴麒瑞, 等. 基于多源异构数据的地质知识图谱构建与应用[J/OL]. 地学前缘, 1-17. [2025-01-15]. https://doi.org/10.13745/j.esf.sf.2024.11.69.
[13]	张前龙, 周永章, 郭兰萱, 等. 找矿知识图谱的智能化应用: 以钦杭成矿带斑岩铜矿为例[J]. 地学前缘, 2024, 31(4): 7-15. DOI
[14]	周永章, 张前龙, 黄永健, 等. 钦杭成矿带斑岩铜矿知识图谱构建及应用展望[J]. 地学前缘, 2021, 28(3): 67-75. DOI
[15]	王成彬, 王明果, 王博, 等. 融合知识图谱的矿产资源定量预测[J]. 地学前缘, 2024, 31(4): 26-36. DOI
[16]	QIU Q J, MA K, LV H R, et al. Construction and application of a knowledge graph for iron deposits using text mining analytics and a deep learning algorithm[J]. Mathematical Geosciences, 2023, 55(3): 423-456.
[17]	WANG B, WU L, XIE Z, et al. Understanding geological reports based on knowledge graphs using a deep learning approach[J]. Computers and Geosciences, 2022, 168: 105229.
[18]	冉一早, 董少春, 王汝成, 等. 铌钽矿床知识图谱的构建及应用实践[J]. 高校地质学报, 2023, 29(3): 359-371.
[19]	YAN Q, XUE L F, LI Y S, et al. Mineral prospectivity mapping integrated with geological map knowledge graph and geochemical data: a case study of gold deposits at Raofeng area, Shaanxi Province[J]. Ore Geology Reviews, 2023, 161: 105651.
[20]	周冠云, 朱成河, 高存山, 等. 大数据框架下金刚石地质成矿信息挖掘及应用[J]. 地球学报, 2023, 44(5): 909-915.
[21]	张春菊, 刘文聪, 张雪英, 等. 基于本体的金矿知识图谱构建方法[J]. 地球信息科学学报, 2023, 25(7): 1269-1281. DOI
[22]	LIN J J, ZHAO Y Z, HUANG W Y, et al. Domain knowledge graph-based research progress of knowledge representation[J]. Neural Computing and Applications, 2021, 33(2): 681-690.
[23]	DUSHYANTHA N, BATAPOLA N, ILANKOON I M S K, et al. The story of rare earth elements (REEs): occurrences, global distribution, genesis, geology, mineralogy and global production[J]. Ore Geology Reviews, 2020, 122: 103521.
[24]	CHEN P, ILTON E S, WANG Z M, et al. Global rare earth element resources: a concise review[J]. Applied Geochemistry, 2024, 175: 106158.
[25]	CHANG Y, KONG L, JIA K J, et al. Chinese named entity recognition method based on BERT[C]// 2021 IEEE International Conference on Data Science and Computer Application (ICDSCA). New York: IEEE, 2021: 294-299.
[26]	SMITH M P, MOORE K, KAVECSÁNSZKI D, et al. From mantle to critical zone: a review of large and giant sized deposits of the rare earth elements[J]. Geoscience Frontiers, 2016, 7(3): 315-334.
[27]	HOSHINO M, SANEMATSU K, WATANABE Y. REE mineralogy and resources[M]// BÜNZLIJ G, PECHARSKYV K. Including actinides. Amsterdam: Elsevier, 2016: 129-291.
[28]	毛景文, 宋世伟, 刘敏, 等. 稀土矿床: 基本特点与全球分布规律[J]. 地质学报, 2022, 96(11): 3675-3697.
[29]	LIU S, FAN H R, WANG Q W, et al. Carbonatite-related delicate REE mineralization processes revealed by fluorocarbonates and monazite: insights from the giant Bayan Obo REE-Nb-Fe deposit, China[J]. Ore Geology Reviews, 2023, 157: 105443.
[30]	LI S L, ZHANG W, CAI J L, et al. Multiple pulses of fluids involved in the formation of carbonatite-related REE deposits as revealed by fluorite[J]. Ore Geology Reviews, 2023, 159: 105546.
[31]	HAO X J, JI Z, LI X H, et al. Construction and application of a knowledge graph[J]. Remote Sensing, 2021, 13(13): 2511.
[32]	刘峤, 李杨, 段宏, 等. 知识图谱构建技术综述[J]. 计算机研究与发展, 2016, 53(3): 582-600.
[33]	程建忠, 车丽萍. 中国稀土资源开采现状及发展趋势[J]. 稀土, 2010, 31(2): 65-69, 85.
[34]	王凯怡, 范宏瑞, 谢奕汉. 白云鄂博碳酸岩墙的稀土和微量元素地球化学及对其成因的启示[J]. 岩石学报, 2002, 18(3): 340-348.
[35]	谢玉玲, 夏加明, 崔凯, 等. 中国碳酸岩型稀土矿床: 时空分布与成矿过程[J]. 科学通报, 2020, 65(33): 3794-3808.
[36]	XU C, KYNICKY J, CHAKHMOURADIAN A R, et al. A case example of the importance of multi-analytical approach in deciphering carbonatite petrogenesis in South Qinling orogen: Miaoya rare-metal deposit, Central China[J]. Lithos, 2015, 227: 107-121.
[37]	杨立强, 刘江涛, 张闯, 等. 哀牢山造山型金成矿系统: 复合造山构造演化与成矿作用初探[J]. 岩石学报, 2010, 26(6): 1723-1739.
[38]	WU Q R, LIU Z H, MIAO T, et al. Chinese mineral exploration named entity recognition for literature mining by fusing multi-features with an enhancement domain pre-training model[J]. Ore Geology Reviews, 2025, 176: 106396.
[39]	CHEN J S, LIN W C, YANG S, et al. Development of anopen-source annotated glaucoma medication dataset from clinical notes in the electronic health record[J]. Translational Vision Science and Technology, 2022, 11(11): 20.
[40]	冯秋睿, 赵汀, 刘超. 基于本体的三稀矿产知识图谱构建[J]. 中国矿业, 2024, 33(4): 79-88.
[41]	叶育鑫, 刘家文, 曾婉馨, 等. 基于本体指导的矿产预测知识图谱构建研究[J]. 地学前缘, 2024, 31(4): 16-25. DOI
[42]	邱芹军, 田苗, 马凯, 等. 区域地质调查文本中文命名实体识别[J]. 地质论评, 2023, 69(4): 1423-1433.
[43]	LIU Y F, WEI S Q, HUANG H J, et al. Naming entity recognition of citrus pests and diseases based on the BERT-BiLSTM-CRF model[J]. Expert Systems with Applications, 2023, 234: 121103.
[44]	LI W, DU Y J, LI X Y, et al. UD_BBC: named entity recognition in social network combined BERT-BiLSTM-CRF with active learning[J]. Engineering Applications of Artificial Intelligence, 2022, 116: 105460.
[45]	QIU Q J, XIE Z, WU L, et al. BiLSTM-CRF for geological named entity recognition from the geoscience literature[J]. Earth Science Informatics, 2019, 12(4): 565-579.
[46]	DEVLIN J, CHANG M W, LEE K, et al. BERT: pre-training of deep bidirectional transformers for language understanding[C]// Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies. Stroudsburg: ACL, 2018, 6: 3498-4195.
[47]	HUANG Z H, XU W, YU K. Bidirectional LSTM-CRF models for sequence tagging[DB/OL]. (2015-08-09) [2025-04-10]. https://doi.org/10.48550/arXiv.1508.01991.
[48]	刘文聪, 张春菊, 汪陈, 等. 基于BiLSTM-CRF的中文地质时间信息抽取[J]. 地球科学进展, 2021, 36(2): 211-220. DOI
[49]	陈忠良, 袁峰, 李晓晖, 等. 基于BERT-BiLSTM-CRF模型的中文岩石描述文本命名实体与关系联合提取[J]. 地质论评, 2022, 68(2): 742-750.
[50]	NIU Z Y, ZHONG G Q, YU H. A review on the attention mechanism of deep learning[J]. Neurocomputing, 2021, 452: 48-62.
[51]	BAHDANAU D, CHO K, BENGIO Y. Neural machine translation by jointly learning to align and translate[DB/OL]. (2016-05-19) [2025-04-10]. https://doi.org/10.48550/arXiv.1409.0473.
[52]	SUTTON C, ROHANIMANESH K, MCCALLUM A. Dynamic conditional random fields: factorized probabilistic models for labeling and segmenting sequence data[J]. Journal of Machine Learning Research, 2007, 8(8): 693-723.
[53]	谢雪景, 谢忠, 马凯, 等. 结合BERT与BiGRU-Attention-CRF模型的地质命名实体识别[J]. 地质通报, 2023, 42(5): 846-855.
[54]	WANG C B, TAN L Q, LI Y J, et al. Ontology-driven relational data mapping for constructing a knowledge graph of porphyry copper deposits[J]. Earth Science Informatics, 2024, 17(3): 2649-2660.
[55]	KARYPIS G, HAN E H, KUMAR V. Chameleon: hierarchical clustering using dynamic modeling[J]. Computer, 1999, 32(8): 68-75.
[56]	张福良, 李政林. 我国独居石资源开发利用现状及政策建议[J]. 现代矿业, 2015, 31(11): 1-4.
[57]	COCHERIE A, ALBAREDE F. An improved U-Th-Pb age calculation for electron microprobe dating of monazite[J]. Geochimica et Cosmochimica Acta, 2001, 65(24): 4509-4522.
[58]	朱永峰, MASSONNE H J. 磷灰石中磁黄铁矿出溶结构的发现[J]. 岩石学报, 2005, 21(2): 405-410.
[59]	李自静, 刘琰. 川西冕宁-德昌REE矿带风化型矿床的矿石类型及成因[J]. 地球科学, 2018, 43(4): 1307-1325.
[60]	侯增谦, 田世洪, 谢玉玲, 等. 川西冕宁-德昌喜马拉雅期稀土元素成矿带: 矿床地质特征与区域成矿模型[J]. 矿床地质, 2008, 27(2): 145-176.
[61]	CHAKHMOURADIAN A R, REGUIR E P, COUËSLAN C, et al. Calcite and dolomite in intrusive carbonatites. II. Trace-element variations[J]. Mineralogy and Petrology, 2016, 110(2): 361-377.
[62]	刘琰, 陈超, 舒小超, 等. 青藏高原东部碳酸岩-正长岩杂岩体型REE矿床成矿模式: 以大陆槽REE矿床为例[J]. 岩石学报, 2017, 33(7): 1978-2000.
[63]	LOUVEL M, ETSCHMANN B, GUAN Q S, et al. Carbonate complexation enhances hydrothermal transport of rare earth elements in alkaline fluids[J]. Nature Communications, 2022, 13: 1456. DOI PMID
[64]	陈衍景, 翟明国, 蒋少涌. 华北大陆边缘造山过程与成矿研究的重要进展和问题[J]. 岩石学报, 2009, 25(11): 2695-2726.
[65]	DUAN Z P, JIANG S Y, SU H M, et al. Textural features and in situ trace element analysis of fluorite from the Wujianfang fluorite deposit, Inner Mongolia (NE China): insights into fluid metasomatism and ore-forming process[J]. Ore Geology Reviews, 2022, 147: 104982.
[66]	谭侯铭睿, 黄小文, 漆亮, 等. 磷灰石化学组成研究进展: 成岩成矿过程示踪及对矿产勘查的指示[J]. 岩石学报, 2022, 38(10): 3067-3084.
[67]	CAMPBELL L S, HENDERSON P. Apatite paragenesis in the Bayan Obo REE-Nb-Fe ore deposit, Inner Mongolia, China[J]. Lithos, 1997, 42(1/2): 89-103.
[68]	李晓春, 展云翔, 范宏瑞, 等. 内蒙古白云鄂博矿床巨量稀土的堆积及再活化历史: 来自矿物微区Sm-Nd同位素的制约[J]. 岩石学报, 2022, 38(10): 2920-2932.
[69]	娄方炬, 顾尚义. 贵州织金寒武纪磷块岩中磷灰石和白云石稀土元素的LA-ICP-MS分析: 对沉积环境和成岩过程的指示意义[J]. 中国稀土学报, 2020, 38(2): 225-239.
[70]	赵芝, 王登红. 氟碳铈矿: 稀土资源的半壁江山[J]. 自然资源科普与文化, 2021(3): 12-15.
[71]	SZUCS A M, STAVROPOULOU A, O’DONNELL C, et al. Reaction pathways toward the formation of bastnäsite: replacement of calcite by rare earth carbonates[J]. Crystal Growth and Design, 2021, 21(1): 512-527.
[72]	舒小超, 刘琰, 李德良, 等. 川西冕宁-德昌稀土矿带霓长岩的地球化学特征及地质意义[J]. 岩石学报, 2019, 35(5): 1372-1388.
[73]	LIU Y, HOU Z Q. A synthesis of mineralization styles with an integrated genetic model of carbonatite-syenite-hosted REE deposits in the Cenozoic Mianning-Dechang REE metallogenic belt, the eastern Tibetan Plateau, southwestern China[J]. Journal of Asian Earth Sciences, 2017, 137: 35-79.
[74]	BENSON E K, WATTS K E. Apatite and monazite geochemistry record magmatic and metasomatic processes in rare earth element mineralization at mountain pass, California[J]. Economic Geology, 2024, 119(7): 1611-1642.
[75]	郑方顺, 宋国学. 铕异常在地质学中的应用[J]. 岩石学报, 2023, 39(9): 2832-2856.
[76]	PACK A, RUSSELL S S, SHELLEY J M G, et al. Geo- and cosmochemistry of the twin elements yttrium and holmium[J]. Geochimica et Cosmochimica Acta, 2007, 71(18): 4592-4608.
[77]	REN Y S, YANG X Y, YANG X M, et al. Mineralogical study on the distribution regularity of niobium in various types of ores in the giant Bayan Obo Fe-REE-Nb deposit[J]. Ore Geology Reviews, 2023, 161: 105602.
[78]	BURNHAM A D, BERRY A J. The effect of oxygen fugacity, melt composition, temperature and pressure on the oxidation state of cerium in silicate melts[J]. Chemical Geology, 2014, 366: 52-60.

属性	实体
矿物	barite、galena、pyrite、aegirine、biotite
岩石	carbonatite、gabbro、basalt、pegmatite、syenite
元素	REE、Ca、Mg、Eu、Sm
年代	Paleogene、Neoproterozoic、Paleozoic、Jurassic、Permian
化学	NaCl、SO₄、SiO₂、TiO₂、Rb/Sr、²⁰⁸Pb/²⁰⁴Pb
矿床	Maoniuping、Dalucao、Lizhuang、Weishan、Bayan Obo
构造	mantle、dike、Mianning-Dechang、vein、Craton、fault

属性	实体
矿物	barite、galena、pyrite、aegirine、biotite
岩石	carbonatite、gabbro、basalt、pegmatite、syenite
元素	REE、Ca、Mg、Eu、Sm
年代	Paleogene、Neoproterozoic、Paleozoic、Jurassic、Permian
化学	NaCl、SO₄、SiO₂、TiO₂、Rb/Sr、²⁰⁸Pb/²⁰⁴Pb
矿床	Maoniuping、Dalucao、Lizhuang、Weishan、Bayan Obo
构造	mantle、dike、Mianning-Dechang、vein、Craton、fault

无用实体	多余符号输出	方法名称	缩写/全称
deep open	diopside (40	ICP-MS	REE/rare earth element
deposit	dolerite,	LA-ICP-MS	LREE/light REE
detail	dolomite <	MC-ICP-MS	MD/Mianning-Dechang
discrete	dolomite δ¹⁸O	EPMA	MNP/Maoniuping

无用实体	多余符号输出	方法名称	缩写/全称
deep open	diopside (40	ICP-MS	REE/rare earth element
deposit	dolerite,	LA-ICP-MS	LREE/light REE
detail	dolomite <	MC-ICP-MS	MD/Mianning-Dechang
discrete	dolomite δ¹⁸O	EPMA	MNP/Maoniuping

节点名称	白云鄂博
图谱名称	calcite			carbonatite			Sr
中间节点数	权重	路径	用时	权重	路径	用时	权重	路径	用时
0	0.192 9	2	<1 s	0.278 2	2	<1 s	0.12	2	<1 s
1	0.338 6	318	<1 s	0.463 3	679	<1 s	0.268 8	267	<1 s
2	0.454 6	18 511	<1 s	0.593 7	26 129	2 s	0.385 8	16 726	<1 s
3	0.540 9	1 328 379	57 s	0.688 6	1 855 502	71 s	0.474 6	1 169 663	49 s
4	0.605 3	86 737 483	3 709 s	0.757 6	118 441 560	4 668 s	0.541 7	77 172 134	3 820 s
5	—	—	—	—	—	—	—	—	—
节点名称	冕宁—德昌
图谱名称	calcite			carbonatite			Sr
中间节点数	权重	路径	用时	权重	路径	用时	权重	路径	用时
0	0.156 9	2	<1 s	0.323 9	2	<1 s	0.132 4	2	<1 s
1	0.292 5	150	<1 s	0.492 4	317	<1 s	0.277 5	167	<1 s
2	0.398 9	4 395	<1 s	0.609 1	7 143	<1 s	0.384 9	4 887	<1 s
3	0.480 4	146 158	5 s	0.693	223 796	7 s	0.464 5	161 150	5 s
4	0.540 8	4 491 471	171 s	0.753 2	6 426 434	216 s	0.523 5	4 900 533	181 s
5	0.584 9	135 771 508	5 188 s	0.796 1	182 967 342	6 537 s	0.567 2	146 940 976	5 549 s