Medical question-answer semantic clustering method based on integrated convolutional encoding

Abstract

The invention discloses a medical question-answer semantic clustering method based on integrated convolutional encoding and relates to the field of machine learning. The method comprises the following steps of collecting question-answer corpuses of medical consultation platform users; selecting convolution kernels; fusing characteristic representations of different convolution kernels; acquiring a final data representation by use of an own encoder; and carrying out medical consultation question-answer semantic clustering. Compared with the conventional deep learning method, different characteristics are extracted with different convolution kernels in the invention, the extracted characteristics are sufficient and diversified, different characteristic merging methods are adopted, the extracted characteristics are subject to fusion representation, thus the method is strong in generalization ability and high in semantic clustering accuracy rate; and based on the method, the own situation of the user can be better understood, the method can assist a doctor in detecting diseases, and great application values are provided for establishment of the automatic medical question-answer system.

Description

technical field [0001] The invention relates to the field of computer artificial intelligence, in particular to the field of machine learning, and in particular to a medical question-and-answer semantic clustering method based on integrated convolution coding. Background technique [0002] With the rapid development of the Internet, people's lifestyles have gradually changed. According to statistical surveys, when ordinary users feel unwell, 90% of them will search for relevant information on the Internet. The Internet is thus changing the medical ecology. In Internet medicine, online disease guidance is a very important and crucial step. As a result, many online disease question-and-answer websites have emerged in the health-related medical field. Patients communicate with doctors and obtain disease-related nursing knowledge by describing their own encounters, detailed conditions, medications, and treatment conditions. In these related disease questions and answers, the...

AI Technical Summary

Problems solved by technology

However, in the medical field, the related technologies of deep learning are not widely used, nor are they reliably applied to the workflow of medical intelligent question answering.

Images