摘要
整理地质资料目录并进行分类统计需要大量专业性工作,为将现代自动化分类技术引入地质资料目录分类工作,在对比了传统机器学习建模和神经网络建模技术的基础上,本文提出了一种基于BERT深度学习模型的地质资料目录自动分类的方法,以此提高地质资料目录分类的准确率。实验结果表明,BERT分类技术的准确率为99%,超过了其他测试模型,且具有较好的易用性和扩展性,能够有效地应用于地质材料目录的自动化分类。
In order to introduce modern automatic classification technology into geological data catalog classification,based on the comparison of traditional machine learning modeling and neural network modeling technology,this paper proposes an automatic classification method of geological data catalog based on BERT deep learning model,so as to improve the accuracy of geological data catalog classification.The experimental results show that the accuracy of BERT classification technology is 99%,which is higher than other test models,and has good ease of use and scalability.It can be effectively applied to the automatic classification of geological material catalogue.
作者
杜晓敏
潘晓
DU Xiaomin;PAN Xiao(China Aero Geophysical Survey and Remote Sensing Center for Natural Resources,Beijing 100083,China;School of Statistics,Renmin University of China,Beijing 100872,China;Development and Research Center,China Geological Survey,Beijing 100037,China)
出处
《中国矿业》
2021年第S02期143-148,共6页
China Mining Magazine
基金
地学信息产品开发与社会化服务地质调查项目资助(编号:DD20190405)
重点地区与特殊地块用途管制监测评价项目资助(编号:DD20211276)。
关键词
BERT
神经网络
文本分类
自然语言处理
地质资料
BERT
neural network
text classification
natural language processing
geological data archives
