Drug recommendation method

Abstract

The invention provides a drug recommendation method. The method comprises steps that the natural language processing technology is utilized to construct a drug knowledge map; the medical record structure technology is utilized to construct a patient portrait based on the patient information, drugs that the patient should not take are filtered according to the drug knowledge map, the patient portrait and the preset recommendation algorithm, and a patient-friendly medication regimen is recommended to a patient. The method is advantaged in that problems that the user information obtained throughthe word-matching-based search method is too small and only a relatively large drug candidate list is recommended are solved, the drugs that the patient should not take are filtered according to the patient portrait and the preset recommendation algorithm, the patient can be recommended for the drugs that are more suitable for the patient, and the situation that a recommended drug is just a contraindication to the user is avoided.

Description

technical field [0001] The invention relates to the field of smart medical care, in particular to a method for drug recommendation. Background technique [0002] The present invention involves main technologies including natural language processing, knowledge graph and recommendation technology. [0003] At present, natural language processing technology, as one of the most important fields of artificial intelligence, has already demonstrated great value in search, advertising, dialogue and other directions. In the medical field, massive medical records, textbooks, papers, questions and answers and other medical data urgently need the development of natural language processing technology. [0004] Knowledge graph is a way of representing knowledge by computer, which has been widely used in search, dialogue and other directions. As a very professional field, medical care contains an extremely large knowledge system. A complete knowledge map can greatly improve the work effi...

AI Technical Summary

Problems solved by technology

Existing similar technologies are used to solve problems in non-medical fields and cannot be directly transplanted to the medical industry

[0006] At present, relevant drug services are mainly searched on the drug e-commerce platform, such as searching for drugs by drug name, disease name, and symptom name. The interaction form is very simple, and there are still the following problems: 1. Word matching based The user information obtained by the search method is too little, and only a large candidate list can be recommended, and the user still cannot know which drug is more suitable for them; 2. There is no user portrait for filtering rational drug use, and there are contraindications for hitting users risk

Images