A Latent Dirichlet Allocation-Based Model for Mining Texts in China's Intelligent Elderly Care Policy

A multitude of information embedded in policy texts reflects the government's policy inclinations, objectives, and impacts. This paper conducts text mining from two dimensions: the policy instruments and implementation topics, establishing a policy text mining model based on LDA, offering a novel approach to policy text mining. By delving into 133 Chinese policy texts on Intelligent Elderly Care, the structural implementation of policy instruments and 10 implementation topics were unearthed. Using multinomial logistic regression, the implementation preferences for different topics' policy instruments were determined, providing insights for stakeholders in the Intelligent Elderly Care policy industry.