Page 21 - 《南京医科大学学报社会科学版》2025年第5期
P. 21
第5期 总第130期 南 ·
南京医科大学学报(社会科学版)京医科大学学报(社会科学版)
2025年10月 Journal of Nanjing Medical University(Social Sciences)of Nanjing Medical University(Social Sciences) · 445
Journal
Text analysis of China’s senior care service policy based on LDA topic
modeling
1,2
LI Xiuping ,ZHANG Jing 1
1. School of Medical Economics and Management,2. Key Laboratory of Data Science and Innovation Development of Traditional
Chinese Medicine,Philosophy and Social Sciences of Anhui Province,Anhui University of Traditional Chinese Medicine,Hefei
230012,China
Abstract: With the deepening of population aging,policies related to senior care services have been
successively introduced,and the focus of policies has been adjusted with the times. This paper selects 44 policy
texts related to senior care service issued from 2015 to 2025 as samples,and applies the Latent Dirichlet Allocation
(LDA)topic model to capture the core themes of the selected policy texts for hot topic identification and
classification. This paper aims to conduct an in ⁃ depth analysis of the development direction and trend of the
Chinese government in the supply of senior care service,revealing the release patterns and theme characteristics of
these policies. 7 key themes were formed in the analysis,with keywords focusing on“health”“community”“society”
“encouragement”and other contents,which were divided into three categories:senior care service supply,senior
care industry and basic senior care service. It is concluded that the senior care service should advance in three key
areas,including establishing a multi⁃level care system and strengthening service supply supervision;diversifying
the industrial form to promote a comprehensive development of senior care;and improving the compatibility
between supply and demand to ensure the long ⁃ term progress of basic senior care service. As a result,these
measures will contribute to the high⁃quality development of China’s senior care system.
Key words: LDA topic model;senior care service;policy text

