Research on the Influence of Video Content Features on User Behaviour


  • Zhiqi Pu Author



User engagement behaviours, Linear regression, Video content features, Social identification theory, Emotional support theory


Research on user engagement behaviours within User-Generated Content (UGC) video platforms is notably scarce, despite previous studies predominantly focusing on user-level information. This study contends that enriched video information holds significant value. Its objective is to provide a profound understanding of the influence mechanisms of video content features on user engagement behaviours within UGC video platforms. Combining exploratory and quantitative methodologies, the research introduces a highly detailed framework for video content features, covering both cognitive and emotional dimensions. The framework encompasses content richness at the video level and emotional features at the user level. Addressing user behaviours, the study encompasses liking, sharing, saving, and tipping, representing users' varied contributions to the platform. The triggers for user behaviours often originate from diverse motivational intentions. The research focuses on a dual perspective, blending user and video viewpoints when examining video content features. Utilizing linear regression equations grounded in social identity theory and emotional support theory, the study explains the role of video content features in triggering user engagement behaviours. Age and gender serve as moderator variables, exploring behavioural disparities between male and female users and across different age groups. Findings indicate that factors triggering user likes and shares primarily stem from the level of interaction in the comment section, while tipping contributions and video saves are influenced by emotional support during viewing. The study also reveals that sadness enhances user participation intentions, while positive emotions in video characters or commenters diminish user engagement intentions. Lastly, the research adopts web crawling through legally accessible interfaces as the primary data collection method, encompassing 435 videos from 25 food category video authors.