Approaches for Quantifying Video Prominence, Narratives, & Discussion: Engagement on COVID-19 Related YouTube Videos.
International Conference on Advances in Social Networks Analysis and Mining(2020)
摘要
Initial scientific studies suggest the spread of extreme content, “the COVID-19 infodemic,” likely plays a crucial role in news spread about the “the COVID-19 pandemic.” In this paper, we quantify the evolution of polarization and engagement in YouTube social networks about public-health interventions for COVID-19. Although YouTube is a major information and news source with high engagement with younger populations, the platform is not widely researched in social network analysis. Discussions about coronavirus on social media can influence how individuals interpret news about the disease and affect their compliance with various non-pharmaceutical interventions. We compare coronavirus video content by identifying three subgroups of public-health intervention-related videos: individual interventions, government interventions, and medical interventions, as well as seven video title narratives. The polarization index measures the level of agreement with the video content using votes: likes and dislikes. The engagement index measures the level of user interaction by comparing views, votes, and number of comments. We observe that over time, engagement for the intervention video subgroups has increased whereas the diffusion for other non-intervention videos has decreased, which suggests that information about COVID-19 interventions has become more popular as the pandemic develops. Additionally, YouTube's search ranking algorithm seems to strongly take into account video polarization as videos that remain prominent in the search rankings have polarization scores 37% lower than videos that are removed from the top ten results. Topicality of video content may also play a role as medical treatment-related videos are the least promoted in the search results amongst the video subgroups despite having low video polarization. Engagement is lowest overall on medical intervention videos, which may be due to vaccine and treatment development as a topic being downgraded quickly from YouTube's search results. We recommend further research into YouTube's search result ranking model to better understand YouTube's role in the spread of news and information about coronavirus and other topics. Despite focusing on the COVID-19 pandemic, the methods for analyzing YouTube videos may be applied to other events or crises.
更多查看译文
关键词
social networks,social media,narratives,polarization,engagement,YouTube,videos,comments,COVID-19,coronavirus
AI 理解论文
溯源树
样例
![](https://originalfileserver.aminer.cn/sys/aminer/pubs/mrt_preview.jpeg)
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要