• 中国计算机学会会刊
  • 中国科技核心期刊
  • 中文核心期刊

Computer Engineering & Science

Previous Articles     Next Articles

Text classification of nursing
adverse events based on CNN-SVM

GE Xiao-wei,LI Kai-xia,CHENG Ming   

  1. (The First Affiliated Hospital of Zhengzhou University,Zhengzhou 450052,China)
     
  • Received:2019-06-14 Revised:2019-08-17 Online:2020-01-25 Published:2020-01-25

Abstract:

The reported contents about the current nursing adverse events are mostly unstructured text data and lack of reasonable and clear classification, so there are many problems such as difficult manual analysis, many human factors, omitted and concealed events, and artificially reduced event levels. This paper proposes a text classification model of Chinese nursing adverse events based on Convolutional Neural Network (CNN) and Support Vector Machine (SVM). A character-level text vocabulary is constructed to vectorize the text, CNN is used to extract the abstract features of the text, and SVM classifier is used to realize the Chinese text classification. Comparative experiments show that the proposed model has better classification effect than the traditional classification models such as SVM and random forest based on TF-IDF (Term Frequency-Inverse Document Frequency) in the text classification of Chinese nursing adverse events.
 

Key words: Chinese text classification, nursing adverse event, CNN-SVM