Exploring YouTube's Recommendation System in the Context of COVID-19 Vaccines: Computational and Comparative Analysis of Video Trajectories.

BACKGROUND: Throughout the COVID-19 pandemic, there has been a concern that social media may contribute to vaccine hesitancy due to the wide availability of antivaccine content on social media platforms. YouTube has stated its commitment to removing content that contains misinformation on vaccination. Nevertheless, such claims are difficult to audit. There is a need for more empirical research to evaluate the actual prevalence of antivaccine sentiment on the internet. OBJECTIVE: This study examines recommendations made by YouTube's algorithms in order to investigate whether the platform may facilitate the spread of antivaccine sentiment on the internet. We assess the prevalence of antivaccine sentiment in recommended videos and evaluate how real-world users' experiences are different from the personalized recommendations obtained by using synthetic data collection methods, which are often used to study YouTube's recommendation systems. METHODS: We trace trajectories from a credible seed video posted by the World Health Organization to antivaccine videos, following only video links suggested by YouTube's recommendation system. First, we gamify the process by asking real-world participants to intentionally find an antivaccine video with as few clicks as possible. Having collected crowdsourced trajectory data from respondents from (1) the World Health Organization and United Nations system (nWHO/UN=33) and (2) Amazon Mechanical Turk (nAMT=80), we next compare the recommendations seen by these users to recommended videos that are obtained from (3) the YouTube application programming interface's RelatedToVideoID parameter (nRTV=40) and (4) from clean browsers without any identifying cookies (nCB=40), which serve as reference points. We develop machine learning methods to classify antivaccine content at scale, enabling us to automatically evaluate 27,074 video recommendations made by YouTube. RESULTS: We found no evidence that YouTube promotes antivaccine content; the average share of antivaccine videos remained well below 6% at all steps in users' recommendation trajectories. However, the watch histories of users significantly affect video recommendations, suggesting that data from the application programming interface or from a clean browser do not offer an accurate picture of the recommendations that real users are seeing. Real users saw slightly more provaccine content as they advanced through their recommendation trajectories, whereas synthetic users were drawn toward irrelevant recommendations as they advanced. Rather than antivaccine content, videos recommended by YouTube are likely to contain health-related content that is not specifically related to vaccination. These videos are usually longer and contain more popular content. CONCLUSIONS: Our findings suggest that the common perception that YouTube's recommendation system acts as a "rabbit hole" may be inaccurate and that YouTube may instead be following a "blockbuster" strategy that attempts to engage users by promoting other content that has been reliably successful across the platform.

Web use remains highly regional even in the age of global platform monopolies.

The Internet, since its inception, has been imagined as a technology that enables information to overcome barriers of language and geography. As a handful of social media platforms now dominate globally, removing most barriers of distribution; this has created unprecedented opportunities for content to gain worldwide traction, regardless of its country of origin. Yet historically with few exceptions, people generally consume content that is from or for their region. Has the Internet or social media really altered this trait? Analyzing the extent of similarities between a hundred countries' web use patterns simultaneously across their most popular websites, and country specific trends from YouTube and Twitter respectively, we find that countries which share borders or where people speak the same languages have the most similar web use patterns. Global social media usage on both YouTube and Twitter is even more heterogeneous and driven to a larger extent by language and geography than global website traffic. Neither does high prevalence of English language speakers in the two countries, nor does one of them being the United States contributes substantially to web use similarity. Global web use remains highly regional. The technical affordances of the Internet alone are thus insufficient to render a cosmopolitan world.

Cascading Crises: Society in the Age of COVID-19.

The tsunami of change triggered by the COVID-19 pandemic has transformed society in a series of cascading crises. Unlike disasters that are more temporarily and spatially bounded, the pandemic has continued to expand across time and space for over a year, leaving an unusually broad range of second-order and third-order harms in its wake. Globally, the unusual conditions of the pandemic-unlike other crises-have impacted almost every facet of our lives. The pandemic has deepened existing inequalities and created new vulnerabilities related to social isolation, incarceration, involuntary exclusion from the labor market, diminished economic opportunity, life-and-death risk in the workplace, and a host of emergent digital, emotional, and economic divides. In tandem, many less advantaged individuals and groups have suffered disproportionate hardship related to the pandemic in the form of fear and anxiety, exposure to misinformation, and the effects of the politicization of the crisis. Many of these phenomena will have a long tail that we are only beginning to understand. Nonetheless, the research also offers evidence of resilience on several fronts including nimble organizational response, emergent communication practices, spontaneous solidarity, and the power of hope. While we do not know what the post COVID-19 world will look like, the scholarship here tells us that the virus has not exhausted society's adaptive potential.