Background Nonpharmaceutical interventions (NPIs) (such as wearing masks and social distancing) have been implemented by governments around the world to slow the spread of COVID-19. To promote public adherence to these regimes, governments need to understand the public perceptions and attitudes toward NPI regimes and the factors that influence them. Twitter data offer a means to capture these insights. Objective The objective of this study is to identify tweets about COVID-19 NPIs in six countries and compare the trends in public perceptions and attitudes toward NPIs across these countries. The aim is to identify factors that influenced public perceptions and attitudes about NPI regimes during the early phases of the COVID-19 pandemic. Methods We analyzed 777,869 English language tweets about COVID-19 NPIs in six countries (Australia, Canada, New Zealand, Ireland, the United Kingdom, and the United States). The relationship between tweet frequencies and case numbers was assessed using a Pearson correlation analysis. Topic modeling was used to isolate tweets about NPIs. A comparative analysis of NPIs between countries was conducted. Results The proportion of NPI-related topics, relative to all topics, varied between countries. The New Zealand data set displayed the greatest attention to NPIs, and the US data set showed the lowest. The relationship between tweet frequencies and case numbers was statistically significant only for Australia ( r =0.837, P <.001) and New Zealand ( r =0.747, P <.001). Topic modeling produced 131 topics related to one of 22 NPIs, grouped into seven NPI categories: Personal Protection (n=15), Social Distancing (n=9), Testing and Tracing (n=10), Gathering Restrictions (n=18), Lockdown (n=42), Travel Restrictions (n=14), and Workplace Closures (n=23). While less restrictive NPIs gained widespread support, more restrictive NPIs were perceived differently across countries. Four characteristics of these regimes were seen to influence public adherence to NPIs: timeliness of implementation, NPI campaign strategies, inconsistent information, and enforcement strategies. Conclusions Twitter offers a means to obtain timely feedback about the public response to COVID-19 NPI regimes. Insights gained from this analysis can support government decision making, implementation, and communication strategies about NPI regimes, as well as encourage further discussion about the management of NPI programs for global health events, such as the COVID-19 pandemic.

When developing topic models, a critical question that should be asked is: How well will this model work in an applied setting? Because standard performance evaluation of topic interpretability uses automated measures modeled on human evaluation tests that are dissimilar to applied usage, these models' generalizability remains in question. In this paper, we probe the issue of validity in topic model evaluation and assess how informative coherence measures are for specialized collections used in an applied setting. Informed by the literature, we propose four understandings of interpretability. We evaluate these using a novel experimental framework reflective of varied applied settings, including human evaluations using open labeling, typical of applied research. These evaluations show that for some specialized collections, standard coherence measures may not inform the most appropriate topic model or the optimal number of topics, and current interpretability performance validation methods are challenged as a means to confirm model quality in the absence of ground truth data.

Recently, research on short text topic models has addressed the challenges of social media datasets. These models are typically evaluated using automated measures. However, recent work suggests that these evaluation measures do not inform whether the topics produced can yield meaningful insights for those examining social media data. Efforts to address this issue, including gauging the alignment between automated and human evaluation tasks, are hampered by a lack of knowledge about how researchers use topic models. Further problems could arise if researchers do not construct topic models optimally or use them in a way that exceeds the models’ limitations. These scenarios threaten the validity of topic model development and the insights produced by researchers employing topic modelling as a methodology. However, there is currently a lack of information about how and why topic models are used in applied research. As such, we performed a systematic literature review of 189 articles where topic modelling was used for social media analysis to understand how and why topic models are used for social media analysis. Our results suggest that the development of topic models is not aligned with the needs of those who use them for social media analysis. We have found that researchers use topic models sub-optimally. There is a lack of methodological support for researchers to build and interpret topics. We offer a set of recommendations for topic model researchers to address these problems and bridge the gap between development and applied research on short text topic models.

BACKGROUND In the absence of a vaccine or curative treatment, non-pharmaceutical interventions regimes have been implemented by governments around the world, to slow the spread of COVID-19. The success of these NPIs has varied between countries and is likely to relate to the degree of uptake and adherence by the community. Understanding public attitudes towards these NPI regimes and the factors that promote public understanding and support for them would provide valuable insight to governments seeking to encourage acceptance of and adherence to these interventions. The analysis of social media offers the opportunity to retrieve these insights. OBJECTIVE The objective of this paper is to describe and compare the public Twitter- based discussion of NPIs implemented during the first four months of the COVID-19 pandemic, across six anglophone countries (Australia, Canada, New Zealand, Republic of Ireland, United Kingdom, and United States). The aim was to determine which NPIs received support from the public and whether there were identifiable elements of public discourse that offered meaningful insight into potential enablers and barriers to community adherence. METHODS We collected 2.5 million tweets related to the COVID-19 pandemic across six countries. These tweets were posted between January 1, 2020, and April 30, 2020. Tweets underwent intensive pre-processing, resulting in the inclusion of 787,691 tweets deemed fit for analysis. A hybrid methodology integrating computational and qualitative analysis was adopted in this study, where the topic model MetaLDA was used to construct topics document sets which were coded and qualitatively analyzed. A total of 94 topics relating to NPIs were identified. A comparative analysis of the public views of NPIs between countries was then conducted. Visualizations and quantitative analysis supported this. RESULTS Our comparison of public discussion of NPIs showed that hand-hygiene was encouraged, and quickly adopted and across all countries surveyed. The spectrum of NPIs aimed at social-distancing required a period of adjustment before being broadly accepted, although, stay-at-home orders were called for in many cases before governments implemented these NPIs. There was contrasting public reaction to restrictions on travel and the closure of businesses. Countries, such as New Zealand, where restrictions were implemented early, showed broad community support and intent to comply. In other countries, such as Canada and the US, there was protracted debate over such restrictions and support for adherence was not as easily interpreted. CONCLUSIONS Attitudes indicative of lowered community adherence to NPIs appear to be rooted in both the complexity of the imposed regime, and the corresponding lack of understanding of the regime in the community. Our study supports the hypothesis that public understanding and responsiveness is supported by consistency, clarity and timeliness of government messaging regarding NPI directives.

No abstract