Metallogenic Prospect Prediction Based On Geological Semantic Model And Machine Learning

Geological and mineral work has accumulated enormous amount of geological data over a long period of time.As a kind of resource,data has attracted more and more attention.How to utilize the existing geological and mineral data and reasonably use new technical methods to obtain the information needed by mineral resources research institutes is an important research direction nowadays.This paper mainly studies the application of information extraction and machine learning based on geological text data in the prediction of metallogenic prospects.By constructing the geological semantics model,the information of text data is extracted and the mineral profile of the study area is determined.Based on the mineral profile and the corresponding geological data,the metallogenic prospect area is predicted by Stochastic Forest algorithm.Taking the western Qinling Mountains of Gansu Province as an example,this study completes the prediction of metallogenic prospects,which provides a reference for further exploration and analysis.It mainly includes the following aspects:(1)Construction of Geological Semantic Model: A basic dictionary of Geology and mineral resources is constructed through research and analysis,and a Geological Semantic Model is constructed by LUIS based on long-term and short-term memory and conditional random field algorithm.(2)Extraction of mineral characteristics: Combining geological and mineral text data and geological semantic model,the corresponding text data meanings are obtained through text data preprocessing,word segmentation,word frequency statistics,etc.,and the related content of ore-controlling elements in mineral species and regional geological characteristics is counted to determine the general situation of mineral resources.(3)Prognosis of metallogenic prospects: According to the general situation of mineral resources,relevant geological data are obtained,and the prediction of metallogenic prospects is realized by stochastic forest machine learning algorithm.(4)Prediction of metallogenic prospects in the West Qinling Mountains of Gansu Province: Taking the West Qinling Mountains of Gansu Province as an example,the feasibility of the method is verified and the metallogenic prospects are predicted.The research shows that the geological semantic model constructed in this paper can extract geological and mineral information from text data,and the prospective metallogenic area can be predicted by stochastic forest algorithm combined with existing geological data.The feasibility of the proposed method is proved by the application of a case study of metallogenic prospecting area prediction in the West Qinling Mountains of Gansu Province.The results provide data reference for further mineral analysis.