Recognition method and recognition system of Chinese medicine named entity based on ancient Chinese medicine literature

Abstract

The invention provides a TCM named entity recognition method and system based on TCM ancient books to solve the problem of TCM named entity recognition. The method performs data cleaning on the basis of obtaining the corpus of ancient Chinese books, and then pre-trains the language model; by sequentially labeling the corpus, a training set for the subsequent model is formed; based on the sequence-labeled model training set, the language model is used as the code Layer, the neural network structure is used as the decoding layer to train the sequence labeling model, so as to perform TCM named entity recognition based on the sequence labeling model. The present invention combines the existing language training model, such as the language model pre-training method bert proposed by Google, which saves the cost of manual labeling based on the small sample training set, improves the recognition effect and accuracy, is easy to operate, and realizes the ancient Chinese medicine literature, In particular, the effective and comprehensive use of medical records in ancient Chinese medicine has laid a good foundation for research in the field of Chinese medicine.

Description

technical field [0001] The invention belongs to the field of information processing and TCM literature, and in particular relates to a TCM named entity recognition method and recognition system based on TCM ancient books and literature. Background technique [0002] Traditional Chinese medicine is extensive and profound. On the one hand, the dissemination of Chinese medicine is through the direct experience of the older generation of medical workers, and on the other hand, it is literature. In the literature of traditional Chinese medicine, a large number of ancient medical records of traditional Chinese medicine are preserved, which contain the experience and diagnosis and treatment methods of many famous old Chinese medicine practitioners. The ancient medical records of traditional Chinese medicine mentioned here refer to the continuous records made by ancient Chinese medicine on the symptoms, causes, prescriptions, and medications of patients when treating diseases. Amon...

AI Technical Summary

Problems solved by technology

However, ancient TCM documents are quite different from other documents in terminology and grammar, and have their own characteristics. The named entity recognition method in the prior art is applied to ancient TCM medical records, and the ideal effect cannot be obtained.

At the same time, there are many difficult grammatical phenomena in the medical records of ancient Chinese medicine, which makes manual labeling difficult and expensive, and further increases the difficulty of named entity recognition in traditional Chinese medicine.

Images