Research on Question Classification Methods in the Medical Field

TL;DR

This paper introduces a new medical question classification dataset and a multi-dimensional neural network model that improves classification accuracy in the medical domain.

Contribution

It provides the first specialized dataset for medical question classification and proposes a novel multi-dimensional feature extraction model.

Findings

Enhanced classification accuracy on medical questions

Effective multi-dimensional feature extraction approach

Addresses data scarcity in medical question classification

Abstract

Question classification is one of the important links in the research of question and answering system. The existing question classification models are more trained on public data sets. At present, there is a lack of question classification data sets in specific fields, especially in the medical field. To make up for this gap, this paper presents a data set for question classification in the medical field. Moreover, this paper proposes a multi-dimensional extraction of the characteristics of the question by combining multiple neural network models, and proposes a question classification model based on multi-dimensional feature extraction. The experimental results show that the proposed method can effectively improve the performance of question classification.