Analysis of stroke disease information needs in online medical community based on LDA model
10.3760/cma.j.cn115682-20211019-04712
- VernacularTitle:基于LDA主题模型的在线医疗社区脑卒中疾病信息需求的分析
- Author:
Xudong HE
1
;
Fangyan YANG
;
Hongmei DUAN
Author Information
1. 北京中医药大学护理学院，北京 102488
- Keywords:
Stroke;
Health education;
Online medical community;
Latent Dirichlet Allocation model
- From:
Chinese Journal of Modern Nursing
2022;28(16):2126-2130
- CountryChina
- Language:Chinese
-
Abstract:
Objective:To understand the needs of online medical community users for stroke disease information, and lay the foundation for providing medical care services that meet the needs of stroke patients.Methods:Data mining was used to examine the question records related to stroke in 3 Chinese online medical communities. The data from August 1, 2020 to July 31, 2021 were collected through the crawler code, and the themes were mined using the Latent Dirichlet Allocation (LDA) model after data cleaning and word segmentation.Results:The subject feature analysis of the online medical community question records showed that the LDA model divided 33 731 question records into 4 aspects and 8 themes of information needs, namely, symptom identification and emergency treatment (symptoms, sudden symptoms and emergency measures) , medical diagnosis and treatment (examination and treatment, medical diagnosis, surgery) , rehabilitation care (daily care, diet and medication) , and patients and caregivers' needs for psychological care knowledge (patients or caregivers' concerns about disease prognosis) .Conclusions:The research results based on LDA model can reflect the information needs of online medical community users with stroke, and provide the development direction and information support for meeting the information needs of stroke specialist care and providing structured health education.
