Tracking COVID-19 Discourse on Twitter in North America: Infodemiology Study Using Topic Modeling and Aspect-Based Sentiment Analysis DOI Creative Commons
Hyeju Jang, Emily Rempel, David R. Roth

et al.

Journal of Medical Internet Research, Journal Year: 2021, Volume and Issue: 23(2), P. e25431 - e25431

Published: Jan. 20, 2021

Background Social media is a rich source where we can learn about people’s reactions to social issues. As COVID-19 has impacted lives, it essential capture how people react public health interventions and understand their concerns. Objective We aim investigate concerns in North America, especially Canada. Methods analyzed COVID-19–related tweets using topic modeling aspect-based sentiment analysis (ABSA), interpreted the results with experts. To generate insights on effectiveness of specific for COVID-19, compared timelines topics discussed timing implementation interventions, synergistically including information aspects our analysis. In addition, further anti-Asian racism, sentiments Asians Canadians. Results Topic identified 20 topics, experts provided interpretations based top-ranked words representative each topic. The interpretation timeline showed that discovered trend are highly related promotions such as physical distancing, border restrictions, handwashing, staying home, face coverings. After training data ABSA human-in-the-loop, obtained 545 aspect terms (eg, “vaccines,” “economy,” “masks”) 60 opinion “infectious” (negative) “professional” (positive), which were used inference key selected by negative overall outbreak, misinformation Asians, positive distancing. Conclusions Analyses natural language processing techniques domain expert involvement produce useful health. This study first analyze Canada comparison United States human-in-the-loop domain-specific ABSA. kind could help agencies well what messages resonating populations who use Twitter, be helpful when designing policy new interventions.

Language: Английский

Unusual purchasing behavior during the early stages of the COVID-19 pandemic: The stimulus-organism-response approach DOI Creative Commons
Samuli Laato, A.K.M. Najmul Islam, Ali Farooq

et al.

Journal of Retailing and Consumer Services, Journal Year: 2020, Volume and Issue: 57, P. 102224 - 102224

Published: July 21, 2020

During the COVID-19 pandemic, unusual consumer behavior, such as hoarding toilet paper, was reported globally. We investigated this behavior when fears of market disruptions started circulating, to capture human in unique situation. Based on stimulus-organism-response (S-O-R) framework, we propose a structural model connecting exposure online information sources (environmental stimuli) two behavioral responses: purchases and voluntary self-isolation. To test proposed model, collected data from 211 Finnish respondents via an survey, carried out analysis using PLS-SEM. found strong link between self-intention self-isolate intention make purchases, providing empirical evidence that directly linked anticipated time spent The results further revealed led increased overload cyberchondria. Information also predictor Perceived severity situation cyberchondria had significant impacts people's voluntarily self-isolate. Future research is needed confirm long-term effects pandemic retail services.

Language: Английский

Citations

740

Tracking Social Media Discourse About the COVID-19 Pandemic: Development of a Public Coronavirus Twitter Data Set DOI Creative Commons
Emily Chen, Kristina Lerman, Emilio Ferrara

et al.

JMIR Public Health and Surveillance, Journal Year: 2020, Volume and Issue: 6(2), P. e19273 - e19273

Published: May 19, 2020

At the time of this writing, novel coronavirus (COVID-19) pandemic outbreak has already put tremendous strain on many countries' citizens, resources and economies around world. Social distancing measures, travel bans, self-quarantines, business closures are changing very fabric societies worldwide. With people forced out public spaces, much conversation about these phenomena now occurs online, e.g., social media platforms like Twitter. In paper, we describe a multilingual Twitter dataset that have been continuously collecting since January 22, 2020. We making our available to research community (https://github.com/echen102/COVID-19-TweetIDs). It is hope contribution will enable study online dynamics in context planetary-scale epidemic unprecedented proportions implications. This could also help track scientific misinformation unverified rumors, or understanding fear panic -- undoubtedly more. Ultimately, may contribute towards enabling informed solutions prescribing targeted policy interventions fight global crisis.

Language: Английский

Citations

740

What social media told us in the time of COVID-19: a scoping review DOI Creative Commons
Shu‐Feng Tsao, Helen Chen,

Therese Tisseverasinghe

et al.

The Lancet Digital Health, Journal Year: 2021, Volume and Issue: 3(3), P. e175 - e194

Published: Jan. 28, 2021

With the onset of COVID-19 pandemic, social media has rapidly become a crucial communication tool for information generation, dissemination, and consumption. In this scoping review, we selected examined peer-reviewed empirical studies relating to during first outbreak starting in November 2019 until May 2020. From an analysis 81 studies, identified five overarching public health themes concerning role online platforms COVID-19. These focused on: (i) surveying attitudes, (ii) identifying infodemics, (iii) assessing mental health, (iv) detecting or predicting cases, (v) analyzing government responses (vi) evaluating quality prevention education videos. Furthermore, our review highlights paucity on application machine learning data related lack documenting real-time surveillance developed with For COVID-19, can play disseminating as well tackling infodemics misinformation.

Language: Английский

Citations

718

Mobile Health Apps on COVID-19 Launched in the Early Days of the Pandemic: Content Analysis and Review DOI Creative Commons
Long Chiau Ming, Noorazrina Untong, Nur Amalina Aliudin

et al.

JMIR mhealth and uhealth, Journal Year: 2020, Volume and Issue: 8(9), P. e19796 - e19796

Published: June 30, 2020

Mobile health (mHealth) app use is a major concern because of the possible dissemination misinformation that could harm users. Particularly, it can be difficult for care professionals to recommend suitable coronavirus disease (COVID-19) education and self-monitoring purposes.This study aims analyze evaluate contents as well features COVID-19 mobile apps. The findings are instrumental in helping identify apps education. results apps' assessment potentially help developers improve or modify their existing designs achieve optimal outcomes.The search mHealth available android-based Play Store iOS-based App was conducted between April 18 May 5, 2020. region where we performed United States, virtual private network used locate access from all countries on Google Store. inclusion criteria were related with no restriction language type. basic comparison requirement free subscription, internet connection, advisory content, size app, ability export data, automated data entry. functionality assessed according knowledge (information COVID-19), tracing mapping cases, home monitoring surveillance, online consultation authority, official run by authorities.Of 223 COVID-19-related apps, only 30 (19.9%) found 28 (44.4%) matched criteria. In assessment, most (10/30, 33.3%) (10/28, 35.7%) scored 4 out 7 points. Meanwhile, outcome (13/30, 43.3%) score 3 compared 35.7%), which 2 (out maximum 5 points). Evaluation functions showed 75.0% (n=36) 48 included do not require 56.3% (n=27) provide symptom advice, 41.7% (n=20) have educational content. terms specific functions, more than half maintained authority information provision. Around 37.5% (n=18) 31.3% (n=15) surveillance respectively, 17% (n=8) equipped an function.Most incorporate infographic while instead providing focused content COVID-19. It important guide users choosing based requirements.

Language: Английский

Citations

565

How Adolescents Use Social Media to Cope with Feelings of Loneliness and Anxiety During COVID-19 Lockdown DOI
Veroline Cauberghe,

Ini Van Wesenbeeck,

Steffi De Jans

et al.

Cyberpsychology Behavior and Social Networking, Journal Year: 2020, Volume and Issue: 24(4), P. 250 - 257

Published: Nov. 13, 2020

Next to physical health problems and economic damage, the coronavirus disease 2019 (COVID-19) pandemic associated lockdown measures taken by governments of many countries are expected cause mental problems. Especially for adolescents, who highly rely on social contacts with peers, prolonged period isolation may have detrimental effects their health. Based mood management theory, current study examines if media beneficial adolescents cope feelings anxiety loneliness during quarantine. A survey among 2,165 (Belgian) (13–19 years old) tested how contributed happiness level, whether different coping strategies (active, relations, humor) mediated these relations. Structural equation modeling revealed that had a higher negative impact adolescents' than anxiety. However, anxious participants indicated use more often actively seek manner adapt situation, lesser extent as way keep in touch friends family. The indirect effect through active was significantly positive. Participants were feeling lonely inclined lacking contact. this strategy not related feelings. Humorous positively happiness, but influenced or To conclude, can be used constructive deal COVID-19

Language: Английский

Citations

472

Public Perception of the COVID-19 Pandemic on Twitter: Sentiment Analysis and Topic Modeling Study DOI Creative Commons
Sakun Boon‐itt, Yukolpat Skunkan

JMIR Public Health and Surveillance, Journal Year: 2020, Volume and Issue: 6(4), P. e21978 - e21978

Published: Oct. 25, 2020

Background COVID-19 is a scientifically and medically novel disease that not fully understood because it has yet to be consistently deeply studied. Among the gaps in research on outbreak, there lack of sufficient infoveillance data. Objective The aim this study was increase understanding public awareness pandemic trends uncover meaningful themes concern posted by Twitter users English language during pandemic. Methods Data mining conducted collect total 107,990 tweets related between December 13 March 9, 2020. analyses included frequency keywords, sentiment analysis, topic modeling identify explore discussion topics over time. A natural processing approach latent Dirichlet allocation algorithm were used most common tweet as well categorize clusters based keyword analysis. Results results indicate three main aspects regarding First, trend spread symptoms can divided into stages. Second, analysis showed people have negative outlook toward COVID-19. Third, modeling, relating outbreak categories: emergency, how control COVID-19, reports Conclusions Sentiment produce useful information about social media alternative perspectives investigate crisis, which created considerable awareness. This shows good communication channel for both These findings help health departments communicate alleviate specific concerns disease.

Language: Английский

Citations

432

Social media influence in the COVID-19 Pandemic DOI Creative Commons
Daniel A. González‐Padilla,

Leonardo Tortolero-Blanco

International braz j urol, Journal Year: 2020, Volume and Issue: 46(suppl 1), P. 120 - 124

Published: July 1, 2020

Never before in human history has it been possible to communicate so quickly during a pandemic, social media platforms have key piece for the dissemination of information; however, there are multiple advantages and disadvantages that must be considered. Responsible use these tools can help disseminate important new information, relevant scientific findings, share diagnostic, treatment, followup protocols, as well compare different approaches globally, removing geographic boundaries first time history. In order responsible useful way, is recommended follow some basic guidelines when sharing information on networks COVID-19 era. this paper, we summarize most influence, advantages, pandemic.

Language: Английский

Citations

427

Digital Health Strategies to Fight COVID-19 Worldwide: Challenges, Recommendations, and a Call for Papers DOI Creative Commons
Guy Fagherazzi, Catherine Goetzinger, Mohammed Rashid

et al.

Journal of Medical Internet Research, Journal Year: 2020, Volume and Issue: 22(6), P. e19284 - e19284

Published: June 4, 2020

The coronavirus disease (COVID-19) pandemic has created an urgent need for coordinated mechanisms to respond the outbreak across health sectors, and digital solutions have been identified as promising approaches address this challenge. This editorial discusses current situation regarding fight COVID-19 well challenges ethical hurdles broad long-term implementation of these solutions. To decrease risk infection, telemedicine used a successful care model in both emergency primary care. Official communication plans should promote facile diverse channels inform people about avoid rumors reduce threats public health. Social media platforms such Twitter Google Trends analyses are highly beneficial trends monitor evolution patients’ symptoms or reaction over time. However, acceptability may face due potential conflicts with users’ cultural, moral, religious backgrounds. Digital tools can provide collective benefits; however, they be intrusive erode individual freedoms leave vulnerable populations behind. demonstrated strong various that tested during crisis. More concerted measures implemented ensure future initiatives will greater impact on epidemic meet most strategic needs ease life who at forefront

Language: Английский

Citations

400

Sentiment Analysis of COVID-19 tweets by Deep Learning Classifiers—A study to show how popularity is affecting accuracy in social media DOI Open Access
Koyel Chakraborty, Surbhi Bhatia, Siddhartha Bhattacharyya

et al.

Applied Soft Computing, Journal Year: 2020, Volume and Issue: 97, P. 106754 - 106754

Published: Sept. 28, 2020

Language: Английский

Citations

364

Twitter Discussions and Emotions About the COVID-19 Pandemic: Machine Learning Approach DOI Creative Commons
Jia Xue, Junxiang Chen, Ran Hu

et al.

Journal of Medical Internet Research, Journal Year: 2020, Volume and Issue: 22(11), P. e20550 - e20550

Published: Oct. 28, 2020

Background It is important to measure the public response COVID-19 pandemic. Twitter an data source for infodemiology studies involving monitoring. Objective The objective of this study examine COVID-19–related discussions, concerns, and sentiments using tweets posted by users. Methods We analyzed 4 million messages related pandemic a list 20 hashtags (eg, “coronavirus,” “COVID-19,” “quarantine”) from March 7 April 21, 2020. used machine learning approach, Latent Dirichlet Allocation (LDA), identify popular unigrams bigrams, salient topics themes, in collected tweets. Results Popular included “virus,” “lockdown,” “quarantine.” bigrams “stay home,” “corona virus,” “social distancing,” “new cases.” identified 13 discussion categorized them into 5 different themes: (1) health measures slow spread COVID-19, (2) social stigma associated with (3) news, cases, deaths, (4) United States, (5) rest world. Across all topics, dominant were anticipation that can be taken, followed mixed feelings trust, anger, fear topics. revealed significant feeling when people discussed new cases deaths compared other Conclusions This showed approaches leveraged study, enabling research evolving discussions during As situation rapidly evolves, several are consistently on Twitter, such as confirmed death rates, preventive measures, authorities government policies, stigma, negative psychological reactions fear). Real-time monitoring assessment concerns could provide useful emergency responses planning. Pandemic-related fear, mental already evident may continue influence trust second wave occurs or there surge current

Language: Английский

Citations

348