Frontiers in Public Health · 2022 · 15 citations · 36 references
Social SciencesHealth BehaviorHealth PromotionHealth CommunicationMedia StudiesSocial Media AnalyticsOnline Health CommunitiesPhysical Activity PromotionSentiment AnalysisSocial MediaSocial Media Health CommunicationSocial Media PromotionCommunity Sentiment AnalysisLive Streaming CommerceLive Streaming CommunicationLive Streaming PlatformsLive TweetingOnline Topic ModelingSocial Media Health Information EvaluationTextual Sentiment Analysis
Objectives: Live-streaming fitness is perceived by the Chinese government as an invaluable means to reduce the prevalence of physical inactivity amid the COVID-19 pandemic. This study aims to investigate the public altitudes of the Chinese people toward live-streaming fitness and provide future health communication strategies on the public promotion of live-streaming fitness accordingly. Methods: This study collected live-streaming fitness-related microblog posts from July 2021 to June 2022 in Weibo, the Chinese equivalent to Twitter. We used the BiLSTM-CNN model to carry out the sentiment analysis, and the structured topic modeling (STM) method to conduct content analysis. Results: This study extracted 114,397 live-streaming fitness-related Weibo posts. Over 80% of the Weibo posts were positive during the period of the study, and over 85% were positive in half of the period. This study finds 8 topics through content analysis, which are fitness during quarantine; cost reduction; online community; celebrity effect; Industry; fitness injuries; live commerce and Zero Covid strategy. Conclusions: It is discovered that the public attitudes toward live-streaming fitness were largely positive. Topics related to celebrity effect (5-11%), fitness injuries (8-16%), live commerce (5-9%) and Zero Covid strategy (16-26%) showed upward trends in negative views of the Chinese people. Specific health communication strategy suggestions are given to target each of the negative topics.
36
Applied Physics Letters · 2000 · 3K citations · Full text
Humanities, Biomedical Informatics, Language Documentation +7
Structural Topic Models for Open‐Ended Survey Responses
Margaret E. Roberts, Brandon Stewart, Dustin Tingley et al. · American Journal of Political Science · 2014 · 1.9K citations
Sentiment Analysis of Comment Texts Based on BiLSTM
Guixian Xu, Yueting Meng, Xiaoyu Qiu et al. · IEEE Access · 2019 · 648 citations · Full text