Drug interaction relationship extraction method and system

Abstract

The invention provides a drug interaction relationship extraction method and system, and the method comprises the steps: inputting a drug specification into a trained machine learning model, and recognizing a drug entity relationship, wherein the machine learning model is built in a mode of semi-supervised learning training for extracting the feature text information in the drug specification; andproviding the drug entity relationship for a pharmacist and a user for reference. According to the drug interaction relationship extraction method and system provided by the embodiment of the invention, a semi-supervised learning mode is adopted, a machine learning model is trained, and drug entity names appearing in a drug specification can be comprehensively judged, so that a result is more accurate.

Description

technical field [0001] The present invention relates to the technical field of information processing, and more specifically, to a method and system for extracting drug interaction relationships. Background technique [0002] People are paying more and more attention to their own health, which invisibly increases the operating load of major pharmacies and makes pharmacists have higher and higher requirements for their own professional knowledge. Under normal circumstances, pharmacists can prescribe appropriate prescriptions to patients based on their own professional knowledge, but due to time and energy limitations, they cannot fully understand each specific drug product. Therefore, the corresponding drug instructions become It is the pharmacist's first reference text on how to administer the drug. However, due to the large amount of text in some drug instructions, it may be difficult for pharmacists to find useful key information in a short period of time, and it is also ...

AI Technical Summary

Problems solved by technology

However, the current general-purpose extraction and processing text information technology has great limitations and is not suitable for the extraction of text information in drug instructions. It has the following shortcomings: 1. The accuracy is not enough. The medical field is a very rigorous field. , which has high requirements on the accuracy of the results, it needs a more accurate model to identify the complex relationship between drug entities, disease entities, food name entities, etc. in drug instructions

2. There is no good solution for the identification and extraction of all the complex entities and their relationships that appear in the drug instructions (such as: drug interactions, indications, contraindications, precautions, etc.)

3. Using pure named entity recognition technology to identify various entity names (including a large number of nested entity names) appearing in drug instructions, lacking the cooperation of a professional pharmacist team, resulting in insufficient recognition accuracy and inconsistent There will be many problems in actual demand

4. The fine-grained division of the text information in the instructions is not enough, which does not meet the actual work needs of pharmacists

5. There is no unified management form for the correctly identified data, and subsequent data utilization cannot be performed, which wastes data

Images