Analysis on health needs of users in HIV/AIDS related online health communities based on text mining technology
Objective To mine the question-and-answer data of users in HIV/AIDS(human immunodeficiency virus/acquired immunodeficiency syndrome)related sections of online health communities,and to to analyze their hidden health needs.Methods Python programming language was used to scrape information from the"patient-patient"and"doctor-patient"online health communities as of August 10,2023.After text preprocessed with Jieba segmentation and CNKI medical thesaurus,t-SNE algorithm for dimensionality reduction clustering and small-batch K-means topic clustering were used to obtain detailed information.A sentiment analysis model was established by TextCNN algorithm.The clustering effectiveness was assessed by silhouette scores.Results A total of 7 online health communities were included,and 258 379 records were collected.The word cloud map represented 40 860 words extracted after segmentation.The top ten high-frequency words were"cheer up","take medicine","hope","know","infection","friends","hospital","partner","feeling"and"like".The t-SNE clustering with a silhouette score of 0.72 detected three thematic categories,namely,health information needs,emotional support needs and social companionship needs.Health information needs encompassed the whole process of disease prevention,diagnosis and treatment.Emotional support needs,as indicated by sentiment analysis chart,showed that positive sentiment support tendency accounted for 24.44%,negative sentiment catharsis tendency accounted for 15.59%and neutral words accounted for 60.97%among the comments from this group.Social companionship needs had the highest proportion and could be further categorized into online and offline friendship.Conclusions Online health community is one of the channels for HIV/AIDS patients and people at high risk of HIV infection to seek social companionship,emotional support and health information.Further optimization of platform-related functions can be carried out to provide a venue for the implementation of user-oriented HIV/AIDS publicity,education,intervention and other prevention and control work.