Research On Viewer Engagement Of Medical And Health Videos In UGC Video Community

The medical and health videos in UGC video community can greatly reduce the threshold and cost for people to obtain information for self-health management,has become an important channel for people to obtain medical and health information.As UGC video community depends on users’ continuous creation and active interaction,stimulating users’ engagement is the key to keep the community alive and play its role in public health management effectively.Therefore,this thesis focuses on the influencing factors and influencing mechanism of viewer engagement intention and behavior of medical and health videos in UGC video community,and on this basis,constructs a prediction model of viewer engagement behavior based on multidimensional attribute features.First of all,this thesis adopts a mixed-methods design that combines qualitative and quantitative research to explore the antecedents and influencing mechanisms of the viewer’s engagement intention of medical and health videos in UGC video community.The results show that the video content quality and source credibility will affect the viewer ’s perceived usefulness and trust through the central route and the peripheral route respectively,and then affect their engagement intention through the change of attitude.However,the viewer’s health concerns will have an impact on the effectiveness of the central route and the peripheral route.Secondly,this thesis uses deep learning algorithm based on BiLSTM-CRF and machine learning classification algorithm based on video image,sound,language style and other dimensional features to identify and evaluate the knowledge and vividness of video,which are two key factors to measure the quality of medical and health video,and explore their influence on viewer engagement behavior.The results show that viewer consuming engagement behavior is mainly influenced by the expectation of title knowledge disclosure,while viewer contributing engagement behavior and creating engagement behavior are related to the direction and size of title-content knowledge deviation.At the same time,the vividness of the video has a positive influence ’on the viewer’s engagement behavior after watching the video,and its influence will be more significant when the knowledge of the video is relatively low.Finally,on the basis of the above empirical research,this thesis further constructs a prediction model of viewer engagement of medical and health videos in UGC video community based on machine learning,and empirically tests the important value of multi-dimensional attribute features of videos in the prediction model.This thesis enriches and expands the research perspective of users’ engagement in social media and virtual community,especially lays a theoretical foundation for the related research of UGC video community and health information context.The conclusion of this thesis has certain guiding significance for the operation and management of UGC video community and the content creation of creators of medical and health videos in UGC video community.