A research team led by Prof. Zhang Nannan from the Xinjiang Institute of Ecology and Geography (XIEG) of the Chinese Academy of Sciences has proposed a deep learning-based method for extracting information related to mineral exploration. This study, published in Ore Geology Reviews, presents an automated solution that efficiently distills essential geological insights from extensive collections of unstructured mineral exploration reports and scientific literature.

Geological reports contain a wealth of data crucial for comprehending geological processes and enhancing the effectiveness of mineral exploration. However, the manual extraction of information from the vast and intricate landscape of geological texts is often an arduous and time-consuming task.

In response to this challenge, the researchers developed a specialized mineral exploration corpus featuring 21 distinct entity types tailored for granitic pegmatite-type lithium deposits. They introduced a multi-feature fusion-based named entity recognition model designed to automate the extraction of mineral exploration information.

"Deep learning methods can rapidly and accurately extract geological knowledge from massive unstructured geological texts," said TAO Jintao, first author of the study.

This study not only delivers a practical solution for automating the extraction of geological information but also enhances the overall understanding of geological knowledge, thereby facilitating more efficient mineral exploration processes.

Article link: https://doi.org/10.1016/j.oregeorev.2024.106367

Workflow of corpus construction. (Image by XIEG)

Contact

LONG Huaping

Xinjiang Institute of Ecology and Geography

E-mail: longhp@ms.xjb.ac.cn

Web: http://english.egi.cas.cn