Intelligent triage method and device based on data enhancement technology, storage equipment and terminal

Abstract

The invention provides an intelligent triage method and device based on a data enhancement technology, a storage device and a terminal, and the method comprises the steps: S10, data obtaining: obtaining a complaint text of a patient, the complaint text comprising an illness state text and a basic information text; S20, preprocessing: preprocessing the chief complaint text; S30, data enhancement: performing data expansion on the preprocessed chief complaint text to obtain a new text set; S40, feature extraction: performing text feature extraction on the new text set to obtain a text feature vector set; S50, model training: training the classifier to obtain a trained classifier model; and S60, triage output: inputting the text feature vector set into a trained classifier model, and taking the category of which the error rate is controlled in a target range as a text classification result by the trained classifier model and outputting the text classification result. The method has the beneficial effect of high accuracy under a relatively small data set, and is suitable for the field of medical informatization.

Description

technical field [0001] The present invention relates to the technical field of medical informatization, in particular to an intelligent triage method, device, storage device and terminal based on data enhancement technology. Background technique [0002] Text classification technology is mainly to analyze and classify text data according to text types and characteristics; for short texts (usually considered to be short texts within 100 words), it has the characteristics of few words, high feature dimension and sparseness, etc., often occurs For online media, social platforms, etc.; for the main complaint text, according to the length distribution, it also belongs to the category of short texts. [0003] In short text classification technology: due to the short number of words and limited features, it is difficult to infer the meaning of sentences with a small amount of information. Under the same algorithm, the classification accuracy is low. [0004] In the current medical...

AI Technical Summary

Problems solved by technology

[0003] In short text classification technology: due to the short number of words and limited features, it is difficult to infer the meaning of sentences with a small amount of information. Under the same algorithm, the classification accuracy is low

[0004] In the current medical environment, patients often need to decide which department to register for when they go to the hospital for treatment; however, because patients usually do not have sufficient professional knowledge, it is inevitable that they cannot choose the correct department for treatment; although Many hospitals arrange medical guides in outpatient clinics, but considering the actual scene of medical visits, it is still far from meeting the needs of patients

[0005] The existing intelligent triage system that provides preliminary triage opinions for patients, although the technical solutions involved have certain triage efficiency, the accuracy rate is still not high enough; the current general deep learning method, its accuracy rate often depends on the data The size and quality of the set, however, due to the sensitivity of medical data and the small amount of data screened out with training value, affect the accuracy rate

Images