数字图书馆论坛2025，Vol.21Issue(5)：40-52,13.DOI:10.3772/j.issn.1673-2286.2025.05.005
融入情感特征的在线健康社区用户画像研究
User Profiling Incorporating Emotional Characteristics in Online Health Communities
摘要
Abstract
Constructing user profiles that incorporate emotional characteristics can identify user needs and provide a basis for optimizing emotional support functions in online health communities.This paper proposes a novel user profiling method incorporating emotional characteristics and an emotional dissemination model based on social network graph structures.Firstly,we construct a social network graph and a user tag feature system by mining information from online health communities.User node features are learned using a multi-head attention graph convolutional network,resulting in fused user node vectors.Subsequently,K-Means++clustering is employed to generate six distinct user profiles.Secondly,an emotional dissemination model is developed by integrating the SEIR epidemiological model to analyze the patterns of emotional dissemination among different users.Empirical research is conducted based on data from the Baidu"Depression Bar"forum.The study systematically characterizes the six identified user profiles and their respective emotional dissemination patterns,and proposes targeted online health community management strategies based on the findings.关键词
在线健康社区/用户画像/情感传播/社交网络图结构/SEIR模型Key words
Online Health Community/User Profiling/Emotional Dissemination/Social Network Graph Structure/SEIR Model分类
社会科学引用本文复制引用
杨磊,李晶晶..融入情感特征的在线健康社区用户画像研究[J].数字图书馆论坛,2025,21(5):40-52,13.基金项目
本研究得到国家社会科学基金后期资助暨优秀博士论文项目"标准化+助推区域产业转型升级的影响机制与政策建议研究"(编号:20FJYB031)、山东省社科规划项目"数字知识产品消费在短视频社交平台的人际扩散机制研究"(编号:22DGLJ26)资助. （编号:20FJYB031）
