Incorporating lexicon and character glyph and morphological features into BiLSTM-CRF for Chinese medical NER

Abstract

Chinese Medical Named Entity Recognition (CMNER) is the basic task of information processing and intelligent medical service in Chinese medical field. In order to make full use of the information of characters and words in the text to improve the effect of CMNER, this paper proposes a recognition model based on character-based BiLSTM-CRF, which integrates lexicon and character features. Firstly, in order to make full use of the information of words and word sequence in the text, the C-ExSoftword method is proposed to integrate the lexicon into the model. According to the similarity of Chinese character glyph and morphologically related forms in the field of Chinese medicine, this paper takes the four-corner coding as the character glyph features of Chinese characters, and extracts the morphological features of each word by using the improved Bidirectional Maximum Matching (BDMM) algorithm. Chinese character glyph features and morphological features are integrated into each character vector. Finally, experiments on real data sets show that the proposed model performs better than the character-based and word-based models.