A New Generation of Artificial Intelligence Algorithms for Mineral Prospectivity Mapping

Zuo, Renguang, Xiong, Yihui, Wang, Ziye, Wang, Jian, and Kreuzer, Oliver P. (2023) A New Generation of Artificial Intelligence Algorithms for Mineral Prospectivity Mapping. Natural Resources Research, 32 (5). pp. 1859-1869.

[img] PDF (Published Version) - Published Version
Restricted to Repository staff only

View at Publisher Website: https://doi.org/10.1007/s11053-023-10237...


Here, we propose a new concept, ‘new generation artificial intelligence (AI) algorithms for mineral prospectivity mapping (MPM)’, which places greater emphasis on interpretability and domain cognitive consistency than the established machine learning (ML) algorithms pertaining to MPM. More specifically, the newly proposed algorithms are designed to (1) allow for the integration of prior geological and expert knowledge into AI models at various stages of the modeling process; (2) offer a degree of transparency about the information transfer process while also improving analysis and evaluation of input features; (3) extract new prospecting information, thereby further enhancing mineral exploration targeting and promoting the advancement of mineral deposit knowledge. We also propose several essential strategies to improve the MPM workflow, including: (1) building a robust conceptual model of the target commodity and deposit type, (2) translating the conceptual model into a practical exploration targeting model, (3) constructing a comprehensive and high-quality geodatabase, and (4) identifying relevant targeting parameters and further integrating them using the newly proposed algorithms. The key motivation behind the development of the new generation AI algorithms for MPM is to improve mineral exploration success rates, a prerequisite to addressing anticipated shortages across a range of critical metallic elements. Drawing from the insights gained in this study, we believe that prioritizing the development of a graph-based AI approach in conjunction with geological expert knowledge would be a valuable direction for future research for MPM.

Item ID: 80424
Item Type: Article (Research - C1)
ISSN: 1573-8981
Keywords: Artificial intelligence, Graph neural networks, Interpretability, Machine learning, Mineral prospectivity mapping, Prior geological knowledge
Copyright Information: © 2023 International Association for Mathematical Geosciences
Date Deposited: 14 Feb 2024 04:15
FoR Codes: 37 EARTH SCIENCES > 3705 Geology > 370508 Resource geoscience @ 100%
SEO Codes: 25 MINERAL RESOURCES (EXCL. ENERGY RESOURCES) > 2503 Mineral exploration > 250399 Mineral exploration not elsewhere classified @ 100%
Downloads: Total: 2
More Statistics

Actions (Repository Staff Only)

Item Control Page Item Control Page