Medical text data-oriented filtering method and system

Abstract

The invention provides a medical text data-oriented filtering method, which comprises the following specific steps of: obtaining a medical record data set S which is not labeled with a medical category label, inputting the medical record data set S into a trained DSSM-C-BiLSTM model, outputting a predicted medical category label Label of the data set S, and performing data filtering through the label Label; the DSSM-C-BiLSTM model training process comprises the following steps: (1) collecting a medical record data set A, performing data preprocessing on the medical record data set A to obtain a data set B, and dividing the data set B into a training set and a test set; (2) constructing a DSSM-C-BiLSTM model, inputting a training set of the data set B into the DSSM-C-BiLSTM model for training and learning, and inputting a test set of the data set B into the trained DSSM-C-BiLSTM model to obtain a medical category label probability; and (3) performing model evaluation index calculation according to the predicted medical category label and the real medical category label, and finishing model training when the model evaluation index meets the condition.

Description

technical field [0001] The invention belongs to the field of natural language processing, and in particular relates to a filtering method and system for medical text data. Background technique [0002] With the rapid development of Internet technology, hospitals store a large amount of medical information and resources, but due to the hysteresis of relevant laws and regulations in the field of health and medical big data, the data has not been fully mined and utilized, and related development has been restricted. In recent years, health and medical big data related industries have been included in the national big data strategic layout, and policies related to medical and health big data have been frequently issued, as well as the development and progress of artificial intelligence technology, using machine learning algorithms to fully mine useful information in medical data become a research hotspot. [0003] The most core, largest and most valuable data in medical data is...

AI Technical Summary

Problems solved by technology

[0004] However, the length of the text sequence of medical records is far greater than the length of the sequence that can be processed by the deep neural network, resulting in the neural network not being able to obtain the entire text sequence information well, and training to obtain the optimal model, so it is necessary to compress and filter the medical data

Images