Chinese Named Entity Recognition in Medicine

Abstract

Entity Recognition (NER) of Chinese material bank names such as text and medical history of social media is a challenging challenge. We describe and appraise character series display methods containing Chinese characters contained in Chinese NER, which develops richer features and improves the performance of the model using multi-task learning, self-interest, and efficient self-awareness training methods. The proposed model is currently the best single model with a rigorous F1 of 90.70% in the electronic history dataset (CCKS-NER 2017).