欢迎访问机床与液压官方网站!

咨询热线:020-32385312 32385313 RSS EMAIL-ALERT
基于BERT的煤矿装备维护知识命名实体识别研究
作者:
作者单位:

作者简介:

通讯作者:

中图分类号:

基金项目:

国家自然科学基金重点资助项目 (51834006);国家自然科学基金面上项目(51875451)


Coal Mine Equipment Maintenance Knowledge Named Entity Recognition Model Based on BERT
Author:
Affiliation:

Fund Project:

  • 摘要
  • |
  • 图/表
  • |
  • 访问统计
  • |
  • 参考文献
  • |
  • 相似文献
  • |
  • 引证文献
  • |
  • 资源附件
  • |
  • 文章评论
    摘要:

    为解决煤矿装备维护知识中语义复杂、实体识别困难的问题,以自建的煤矿装备维护知识语料库为研究对象,提出一种基于BERT的煤矿装备维护知识命名实体识别方法。利用BERT获取词的语义、归属及位置信息,增强词向量的语义表征能力;然后将词向量序列输入BiLSTM层,获取上下文信息并提取长距离特征;最后利用CRF对序列标记进行合法性约束;并对模型进行超参数优化,减少特征损失并提高学习效率。实验结果表明:所提方法准确率、召回率和F1值显著提升,分别达到90.32%、93.82%、91.54%,证明该模型有效改善了煤矿装备维护实体中一词多义及重叠实体识别困难问题。

    Abstract:

    In order to solve the problems of complex semantics and difficult entity recognition in coal mine equipment maintenance knowledge,a named entity recognition method of coal mine equipment maintenance knowledge based on BERT was proposed,taking the self-built coal mine equipment maintenance knowledge corpus as the research object.BERT was used to obtain the semantic,attribution and location information of words to enhance the semantic representation ability of word vector.Then,word vector sequences were input into BiLSTM layer to obtain context information and extract long-distance features.Finally,CRF was used to constrain the validity of sequence markers.The model was optimized by hyperparameter to reduce feature loss and improve learning efficiency.The experimental results show that the accuracy rate,recall rate and F1 value of the proposed method are significantly improved,reaching 90.32%,93.82% and 91.54% respectively,which proves that the model can effectively improve the difficult problems of polysemy and overlapping entity recognition in coal mine equipment maintenance entities.

    参考文献
    相似文献
    引证文献
引用本文

曹现刚,吴可昕,张梦园,段雍,李鹏飞.基于BERT的煤矿装备维护知识命名实体识别研究[J].机床与液压,2023,51(9):103-108.
CAO Xiangang, WU Kexin, ZHANG Mengyuan, DUAN Yong, LI Pengfei. Coal Mine Equipment Maintenance Knowledge Named Entity Recognition Model Based on BERT[J]. Machine Tool & Hydraulics,2023,51(9):103-108

复制
分享
文章指标
  • 点击次数:
  • 下载次数:
  • HTML阅读次数:
  • 引用次数:
历史
  • 收稿日期:
  • 最后修改日期:
  • 录用日期:
  • 在线发布日期: 2023-05-29
  • 出版日期: 2023-05-15