Fig. 4From: A deep learning model incorporating part of speech and self-matching attention for named entity recognition of Chinese electronic medical recordsPerformance comparison of different models on the five types of entitiesBack to article page