摘要
针对数控(computer numerical control,CNC)机床故障领域命名实体识别方法中存在实体规范不足及有效实体识别模型缺乏等问题,制定了领域内实体标注策略,提出了一种基于双向转换编码器(bidirectional encoder representations from transformers,BERT)的数控机床故障领域命名实体识别方法。采用BERT编码层预训练,将生成向量输入到双向长短期记忆网络(bidirectional long short-term memory,BiLSTM)交互层以提取上下文特征,最终通过条件随机域(conditional random field,CRF)推理层输出预测标签。实验结果表明,BERT-BiLSTM-CRF模型在数控机床故障领域更具优势,与现有模型相比,F_(1)提升大于1.85%。
Aiming at the problems of insufficient entity specification and lack of effective entity recognition models in the named entity recognition method in the field of computer numerical control(CNC)machine tool faults,an entity labeling strategy in the field was formulated,and a named entity recognition method based on bidirectional encoder representations from transformers(BERT)in the field of CNC machine tool faults was proposed.The BERT coding layer was pre-trained,and the generated vector was input to the bidirectional long short-term memory(BiLSTM)interaction layer to extract contextual features,and finally the predicted label was output through the conditional random field(CRF)inference layer.Experimental results show that BERT-BiLSTM-CRF model has more advantages in the field of CNC machine tool failures.Compared with the existing models,F_(1) value increased by more than 1.85%.
作者
褚燕华
蒋文
王丽颖
张晓琳
王乾龙
CHU Yan-hua;JIANG Wen;WANG Li-ying;ZHANG Xiao-lin;WANG Qian-long(School of Information Engineering, Inner Mongolia University of Science and Technology, Baotou 014010, China)
出处
《科学技术与工程》
北大核心
2022年第14期5737-5743,共7页
Science Technology and Engineering
基金
国家自然科学基金(61562065)。
关键词
命名实体识别
数控机床故障领域
双向转换编码器
named entity recognition
CNC machine tool fault field
bidirectional encoder representations from transformers
