A Method for Extracting Clinical Diagnosis and Treatment Knowledge for Traditional Chinese Medicine Literature

To realize automatic extraction of rich clinical diagnosis and treatment knowledge from traditional Chinese medicine (TCM) literature, a method using information extraction technology to achieve automatic acquisition of clinical diagnosis and treatment knowledge in TCM is proposed. First, the entity types of TCM clinical diagnosis and treatment were defined, and the corpus base of TCM diagnosis and treatment was constructed. Then, the literature sample set was labelled based on the corpus, several commonly used information extraction models were selected to train the sample set, and the most appropriate models and training parameters were selected to extract clinical diagnosis and treatment knowledge. Finally, a case study of breast cancer was carried out, and the implementation scheme of diagnosis and treatment knowledge extraction based on the literature is described in detail. The experimental results show that the UIE model has the best information extraction effect, and its F1 value is 89.69%. Data mining was carried out on the extracted diagnosis and treatment knowledge, and it was found that the basic principles of drug use were qi and blood circulation and liver and spleen strengthening, and the main drugs were tonifying deficiency and clearing heat. This method provides an effective technical route for the automatic acquisition of TCM clinical diagnosis and treatment knowledge, which can help clinical researchers quickly and accurately obtain diagnosis and treatment knowledge in the literature and has practical value for TCM clinical research.