A Novel Chinese Traditional Medicine Prescription Recommendation System based on Knowledge Graph

As the traditional Chinese medical approach, Chinese medicine plays an extremely important role in the field of medical treatment. With the development of computers, people tend to acquire medical knowledge from the Internet in daily life. Because of the complexity of online Chinese medicine knowledge, nowadays there is no good way to organize the existing knowledge to provide convenience for doctors and patients. This paper introduces the study on the recommendation system based on the Knowledge Graph (KG). Firstly, it conducts extraction of entities such as Traditional Chinese Medicine (TCM) diseases, prescription, Chinese herbal medicine, symptoms, etc. Secondly, it transforms KG into vector space using Node2vec. At last, based on the similarities between vectors the provided system recommends prescription by adopting the diagnostic process of Traditional Chinese Medicine. The result shows that the Hit Ratio (HR) of the recommend system is as high as 80%.