Sub-national longitudinal and geospatial analysis of COVID-19 tweets

Using geospatial social media data for infectious disease studies: a systematic review

Geospatial social media (GSM) data has been increasingly used in public health due to its rich, timely, and accessible spatial information, particularly in infectious disease research. This review synthesized 86 research articles that use GSM data in infectious diseases published between December 2013 and March 2022. These articles cover 12 infectious disease types ranging from respiratory infectious diseases to sexually transmitted diseases with spatial levels varying from the neighborhood, county, state, and country. We categorized these studies into three major infectious disease research domains: surveillance, explanation, and prediction. With the assistance of advanced statistical and spatial methods, GSM data has been widely and deeply applied to these domains, particularly in surveillance and explanation domains. We further identified four knowledge gaps in terms of contextual information use, application scopes, spatiotemporal dimension, and data limitations and proposed innovation opportunities for future research. Our findings will contribute to a better understanding of using GSM data in infectious diseases studies and provide insights into strategies for using GSM data more effectively in future research.

Analyzing Discussions Around Rural Health on Twitter During the COVID-19 Pandemic: Social Network Analysis of Twitter Data

Background

Individuals from rural areas are increasingly using social media as a means of communication, receiving information, or actively complaining of inequalities and injustices.

Objective

The aim of our study is to analyze conversations about rural health taking place on Twitter during a particular phase of the COVID-19 pandemic.

Methods

This study captured 57 days' worth of Twitter data related to rural health from June to August 2021, using English-language keywords. The study used social network analysis and natural language processing to analyze the data.

Results

It was found that Twitter served as a fruitful platform to raise awareness of problems faced by users living in rural areas. Overall, Twitter was used in rural areas to express complaints, debate, and share information.

Conclusions

Twitter could be leveraged as a powerful social listening tool for individuals and organizations that want to gain insight into popular narratives around rural health.

Public Figure Vaccination Rhetoric and Vaccine Hesitancy: Retrospective Twitter Analysis

Background

Social media has emerged as a critical mass communication tool, with both health information and misinformation now spread widely on the web. Prior to the COVID-19 pandemic, some public figures promulgated anti-vaccine attitudes, which spread widely on social media platforms. Although anti-vaccine sentiment has pervaded social media throughout the COVID-19 pandemic, it is unclear to what extent interest in public figures is generating anti-vaccine discourse.

Objective

We examined Twitter messages that included anti-vaccination hashtags and mentions of public figures to assess the connection between interest in these individuals and the possible spread of anti-vaccine messages.

Methods

We used a data set of COVID-19-related Twitter posts collected from the public streaming application programming interface from March to October 2020 and filtered it for anti-vaccination hashtags "antivaxxing," "antivaxx," "antivaxxers," "antivax," "anti-vaxxer," "discredit," "undermine," "confidence," and "immune." Next, we applied the Biterm Topic model (BTM) to output topic clusters associated with the entire corpus. Topic clusters were manually screened by examining the top 10 posts most highly correlated in each of the 20 clusters, from which we identified 5 clusters most relevant to public figures and vaccination attitudes. We extracted all messages from these clusters and conducted inductive content analysis to characterize the discourse.

Results

Our keyword search yielded 118,971 Twitter posts after duplicates were removed, and subsequently, we applied BTM to parse these data into 20 clusters. After removing retweets, we manually screened the top 10 tweets associated with each cluster (200 messages) to identify clusters associated with public figures. Extraction of these clusters yielded 768 posts for inductive analysis. Most messages were either pro-vaccination (n=329, 43%) or neutral about vaccination (n=425, 55%), with only 2% (14/768) including anti-vaccination messages. Three main themes emerged: (1) anti-vaccination accusation, in which the message accused the public figure of holding anti-vaccination beliefs; (2) using "anti-vax" as an epithet; and (3) stating or implying the negative public health impact of anti-vaccination discourse.

Conclusions

Most discussions surrounding public figures in common hashtags labelled as "anti-vax" did not reflect anti-vaccination beliefs. We observed that public figures with known anti-vaccination beliefs face scorn and ridicule on Twitter. Accusing public figures of anti-vaccination attitudes is a means of insulting and discrediting the public figure rather than discrediting vaccines. The majority of posts in our sample condemned public figures expressing anti-vax beliefs by undermining their influence, insulting them, or expressing concerns over public health ramifications. This points to a complex information ecosystem, where anti-vax sentiment may not reside in common anti-vax-related keywords or hashtags, necessitating further assessment of the influence that public figures have on this discourse.

Racially/ethnically stratified COVID-19 tweets are associated with COVID-19 cases and deaths

BACKGROUND

The Coronavirus Disease 2019 (COVID-19) pandemic exacerbated existing racial/ethnic health disparities in the United States (U.S.). Monitoring nationwide Twitter conversations about COVID-19 and race/ethnicity could shed light on the impact of the pandemic on the racial/ethnic minorities and help address health disparities.

OBJECTIVE

This paper aims to examine the association between COVID-19 tweet volume and COVID-19 cases and deaths, stratified by race/ethnicity, in the early onset of the pandemic.

METHODS

This cross-sectional study used geo-tagged COVID-19 tweets from within the U.S. posted in April 2020 on Twitter to examine the association between tweet volume, COVID-19 surveillance data (total cases and deaths in April), and population size. The studied time frame was limited to April 2020 because April was the earliest month when COVID-19 surveillance data on racial/ethnic groups was collected. Racially/ethnically stratified tweets were extracted using racial/ethnic group-related keywords (Asian, Black, Latino, and White) from COVID-19 tweets. Racially/ethnically stratified tweets, COVID-19 cases, and deaths were mapped to reveal their spatial distribution patterns. The ordinary least squares (OLS) regression model was applied to each stratified dataset.

RESULTS

The racially/ethnically stratified tweet volume was associated with surveillance data. Specifically, the increase of one Asian tweet was correlated to 288 Asian cases (p<0.05) and 93.4 Asian deaths (p<0.05); the increase of one Black tweet was linked to 47.6 Black deaths (p<0.05); the increase of one Latino tweets was linked to 719 Latino deaths (p<0.05); and the increase of one White tweet was linked to 60.2 White deaths (p<0.05).

CONCLUSIONS

Using racially/ethnically stratified Twitter data as a surveillance indicator could inform epidemiologic trends to help estimate future surges of COVID-19 cases and potential future outbreaks of a pandemic among racial/ethnic groups.

Spatiotemporal sentiment variation analysis of geotagged COVID-19 tweets from India using a hybrid deep learning model

India is a hotspot of the COVID-19 crisis. During the first wave, several lockdowns (L) and gradual unlock (UL) phases were implemented by the government of India (GOI) to curb the virus spread. These phases witnessed many challenges and various day-to-day developments such as virus spread and resource management. Twitter, a social media platform, was extensively used by citizens to react to these events and related topics that varied temporally and geographically. Analyzing these variations can be a potent tool for informed decision-making. This paper attempts to capture these spatiotemporal variations of citizen reactions by predicting and analyzing the sentiments of geotagged tweets during L and UL phases. Various sentiment analysis based studies on the related subject have been done; however, its integration with location intelligence for decision making remains a research gap. The sentiments were predicted through a proposed hybrid Deep Learning (DL) model which leverages the strengths of BiLSTM and CNN model classes. The model was trained on a freely available Sentiment140 dataset and was tested over manually annotated COVID-19 related tweets from India. The model classified the tweets with high accuracy of around 90%, and analysis of geotagged tweets during L and UL phases reveal significant geographical variations. The findings as a decision support system can aid in analyzing citizen reactions toward the resources and events during an ongoing pandemic. The system can have various applications such as resource planning, crowd management, policy formulation, vaccination, prompt response, etc.

Temporal Variations and Spatial Disparities in Public Sentiment Toward COVID-19 and Preventive Practices in the United States: Infodemiology Study of Tweets

Background

During the COVID-19 pandemic, US public health authorities and county, state, and federal governments recommended or ordered certain preventative practices, such as wearing masks, to reduce the spread of the disease. However, individuals had divergent reactions to these preventive practices.

Objective

The purpose of this study was to understand the variations in public sentiment toward COVID-19 and the recommended or ordered preventive practices from the temporal and spatial perspectives, as well as how the variations in public sentiment are related to geographical and socioeconomic factors.

Methods

The authors leveraged machine learning methods to investigate public sentiment polarity in COVID-19–related tweets from January 21, 2020 to June 12, 2020. The study measured the temporal variations and spatial disparities in public sentiment toward both general COVID-19 topics and preventive practices in the United States.

Results

In the temporal analysis, we found a 4-stage pattern from high negative sentiment in the initial stage to decreasing and low negative sentiment in the second and third stages, to the rebound and increase in negative sentiment in the last stage. We also identified that public sentiment to preventive practices was significantly different in urban and rural areas, while poverty rate and unemployment rate were positively associated with negative sentiment to COVID-19 issues.

Conclusions

The differences between public sentiment toward COVID-19 and the preventive practices imply that actions need to be taken to manage the initial and rebound stages in future pandemics. The urban and rural differences should be considered in terms of the communication strategies and decision making during a pandemic. This research also presents a framework to investigate time-sensitive public sentiment at the county and state levels, which could guide local and state governments and regional communities in making decisions and developing policies in crises.

A longitudinal and geospatial analysis of COVID-19 tweets during the early outbreak period in the United States

INTRODUCTION

Early reports of COVID-19 cases and deaths may not accurately convey community-level concern about the pandemic during early stages, particularly in the United States where testing capacity was initially limited. Social media interaction may elucidate public reaction and communication dynamics about COVID-19 in this critical period, during which communities may have formulated initial conceptions about the perceived severity of the pandemic.

METHODS

Tweets were collected from the Twitter public API stream filtered for keywords related to COVID-19. Using a pre-existing training set, a support vector machine (SVM) classifier was used to obtain a larger set of geocoded tweets with characteristics of user self-reporting COVID-19 symptoms, concerns, and experiences. We then assessed the longitudinal relationship between identified tweets and the number of officially reported COVID-19 cases using linear and exponential regression at the U.S. county level. Changes in tweets that included geospatial clustering were also assessed for the top five most populous U.S. cities.

RESULTS

From an initial dataset of 60 million tweets, we analyzed 459,937 tweets that contained COVID-19-related keywords that were also geolocated to U.S. counties. We observed an increasing number of tweets throughout the study period, although there was variation between city centers and residential areas. Tweets identified as COVID-19 symptoms or concerns appeared to be more predictive of active COVID-19 cases as temporal distance increased.

CONCLUSION

Results from this study suggest that social media communication dynamics during the early stages of a global pandemic may exhibit a number of geospatial-specific variations among different communities and that targeted pandemic communication is warranted. User engagement on COVID-19 topics may also be predictive of future confirmed case counts, though further studies to validate these findings are needed.