摘要
文本分类是自然语言领域一个重要的研究方向和技术核心,一直受到研究者的热切关注。在医学领域,中医源远流长,在人类历史发展中发挥着不可磨灭的作用。中医语言包含了大量中医领域术语,且多为表述严谨和富含辩证思维的古文,上下文词语关联性较强,且大多是结构化、半结构化或非结构化数据的形式,这些特点给中医病案的智能分析分类造成了很大地困难。该文基于注意力机制的深度学习模型Bert模型实现中医深层全局语义的特征表示,并进行中医临床文本的分类研究。最后通过对中医临床文本分类实验的验证,该模型达到了非常可观的分类效果。
Text classification is an important research direction and technical core in the field of natural language,and it has always received eager attention from researchers.In the field of medicine,Chinese medicine has a long history and has played an indelible role in the development of human history.The language of Chinese medicine contains a large number of terms in the field of Chi⁃nese medicine,and most of them are ancient texts with rigorous expressions and rich dialectical thinking.The contextual words are strongly related,and most of them are in the form of structured,semi-structured or unstructured data.These characteristics give Chinese medicine The intelligent analysis and classification of medical records caused great difficulties.In this paper,the Bert model,a deep learning model based on the attention mechanism,realizes the feature representation of the deep global semantics of TCM,and conducts the classification research of TCM clinical text.Finally,through the verification of the TCM clinical text classi⁃fication experiment,the model achieved a very considerable classification effect.
作者
王培
王亚文
卢苗苗
WANG Pei;WANG Ya-wen;LU Miao-miao(North China University of Technology,Tangshan 063000,China)
出处
《电脑知识与技术》
2021年第27期13-14,20,共3页
Computer Knowledge and Technology
基金
河北省“三三三人才工程”培养经费项目(项目编号:A201803082)。
关键词
文本分类
深度学习
中医文本
注意力机制
text categorization
deep learning
TCM text
attention mechanism