基于LDA模型的老年人生活满意度主题挖掘与文本实证分析

Theme Mining and Text Empirical Analysis of Elderly Life Satisfaction Based on LDA Model

通过对老年人生活满意度文献进行分析,明确老年人生活满意度研究的主要方向,为提高老年人生活满意度提供科学建议。检索爬取了CNKI(中国知网)、万方、维普中关于老年人生活满意度领域的文献,将文献摘要部分作为研究对象,采用LDA模型挖掘文本主题,分析主题演进趋势,使用回归方程识别冷、热点主题。去除不相关文献后纳入文献402篇,使用潜在狄利克雷分配(LDA)模型挖掘出8个主题,其中利用回归方程识别热点主题1个(家庭关系与情绪健康),冷点主题1个(再教育状况与生活满意度),发现主题研究趋势进行可视化展示。建议未来的老年人生活满意度研究应更加重视老年人教育、知识、能力的研究,从而提高老年人生活质量和满意度。

To analyze the literature on elderly life satisfaction and clarify the main direction of research on elderly life satisfaction,in order to provide scientific suggestions for improving elderly life satisfaction.Methods The CNKI(China Knowledge Network),Wanfang,and Weipu literatures on life satisfaction of the elderly were retrieved,and the literature summary was taken as the research object.The LDA model was used to mine text topics,analyze the theme evolution trend,and use regression equations to identify cold and hot topics.After removing irrelevant literature,402 articles were included and 8 themes were identified using the latent Dirichlet distribution(LDA)model.Among them,regression equations were used to identify 1 hot topic(family relationships and emotional health)and 1 cold topic(re education status and life satisfaction),and the research trends of the themes were visualized.The conclusion suggests that future research on elderly life satisfaction should pay more attention to the education,knowledge,and abilities of the elderly,in order to improve their quality of life and satisfaction.