A fairness assessment of mobility-based COVID-19 case prediction models.

In light of the outbreak of COVID-19, analyzing and measuring human mobility has become increasingly important. A wide range of studies have explored spatiotemporal trends over time, examined associations with other variables, evaluated non-pharmacologic interventions (NPIs), and predicted or simulated COVID-19 spread using mobility data. Despite the benefits of publicly available mobility data, a key question remains unanswered: are models using mobility data performing equitably across demographic groups? We hypothesize that bias in the mobility data used to train the predictive models might lead to unfairly less accurate predictions for certain demographic groups. To test our hypothesis, we applied two mobility-based COVID infection prediction models at the county level in the United States using SafeGraph data, and correlated model performance with sociodemographic traits. Findings revealed that there is a systematic bias in models' performance toward certain demographic characteristics. Specifically, the models tend to favor large, highly educated, wealthy, young, and urban counties. We hypothesize that the mobility data currently used by many predictive models tends to capture less information about older, poorer, less educated and people from rural regions, which in turn negatively impacts the accuracy of the COVID-19 prediction in these areas. Ultimately, this study points to the need of improved data collection and sampling approaches that allow for an accurate representation of the mobility patterns across demographic groups.

Anatomy into the battle of supporting or opposing reopening amid the COVID-19 pandemic on Twitter: A temporal and spatial analysis.

Reopening amid the COVID-19 pandemic has triggered a battle on social media. The supporters perceived that the lockdown policy could damage the economy and exacerbate social inequality. By contrast, the opponents believed it was necessary to contain the spread and ensure a safe environment for recovery. Anatomy into the battle is of importance to address public concerns, beliefs, and values, thereby enabling policymakers to determine the appropriate solutions to implement reopening policy. To this end, we investigated over 1.5 million related Twitter postings from April 17 to May 30, 2020. With the aid of natural language processing (NLP) techniques and machine learning classifiers, we classified each tweet into either a "supporting" or "opposing" class and then investigated the public perception from temporal and spatial perspectives. From the temporal dimension, we found that both political and scientific news that were extensively discussed on Twitter led to the perception of opposing reopening. Further, being the first mover with full reopen adversely affected the public reaction to reopening policy, while being the follower or late mover resulted in positive responses. From the spatial dimension, the correlation and regression analyses suggest that the state-level perception was very likely to be associated with political affiliation and health value.