A chronological and geographical analysis of personal reports of COVID-19 on Twitter from the UK

Golder, Su; Klein, Ari Z.; Magge, Arjun; O’Connor, Karen; Cai, Haitao; Weissenbacher, Davy; Gonzalez-Hernandez, Graciela

doi:10.1177/20552076221097508

Cited by 7 publications

(3 citation statements)

References 25 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In the random sample of tweets in our study, none included personal experiences of climate change. This stands in contrast to the public's wider use of Twitter, which has been found to provide a safe and anonymous space for sharing health experiences and connecting with others with similar experiences [40,41].…”

Section: Discussionmentioning

confidence: 86%

Social media engagement in health and climate change: an exploratory analysis of Twitter

Golder,

Graham

2024

Environ. Res.: Health

Self Cite

View full text Add to dashboard Cite

Social media are increasingly used by the public to share information and opinions. This study explores social media engagement in health and climate change through an analysis of English-language posts on Twitter, one of the most widely-used platforms. We searched Twitter from 3rd April 2023 to 11th May 2023 for posts related to climate change using climate-related textwords and hashtags; we then used health keywords (‘health’, ‘wellbeing’, ‘illness’, ‘illnesses’, ‘disease’, ‘death’) to identify posts related to health. Focusing on posts from general public users, we investigated the proportion of climate change posts referring to health and, for a random sample of these tweets, undertook a content analysis to identify the ways in which climate change and health were represented. The content analysis drew on media research on ‘framing’, a selective process through which particular aspects of an issue – for example, its causes, impacts and solutions – are highlighted. 668,810 posts related to climate change were posted during the study period. Health-related textwords were included in 2.3% (15,434) of these posts. The content analysis pointed to two divergent frames. The first frame represents climate change as real, with real effects on people’s health. The second frame portrays climate change as a hoax, with hoax-generated health effects. While the ‘reality’ frame does not engage with the hoax frame, the latter provides an insistent counter-narrative that questions trust in mainstream science and government policy. Neither frame engages with people’s experiences of health and climate change. In conclusion, our study points to low levels of engagement in health in a key forum for public discussions about climate change. It also asks whether the failure of the ‘reality’ frame to engage either with people’s lived experiences or with hoax framings may be contributing to a polarised debate about climate change and health and hindering consensus-building.

show abstract

Section: Discussionmentioning

confidence: 86%

Social media engagement in health and climate change: an exploratory analysis of Twitter

Golder,

Graham

2024

Environ. Res.: Health

Self Cite

View full text Add to dashboard Cite

show abstract

“…To the best of our knowledge, our study is the first to examine the Twitter users’ reports of health experiences that could be indicative of influenza and COVID-19 infections in Canada. Previous studies have assessed Twitter users’ self-reports of plausible COVID-19 exposure or infections for tracking COVID-19 spread in the US and UK ( 20 , 21 ), and other Canadian studies using COVID-19 related Twitter data are focused on analyzing organizational tweets ( 41 ), COVID-19 tweets posted by federal members of parliament ( 42 ), or more general sentiment analysis using Twitter emojis ( 43 ).…”

Section: Discussionmentioning

confidence: 99%

“…This could offer useful complementary information for case detection. Previous studies have explored voluntary self-reporting of potential exposure to COVID-19 by Twitter users in the US and UK ( 20 , 21 ), but there is limited research on assessing user-level personal reports of influenza and COVID-19 in Canada. The first objective of our study is to assess whether English-speaking Canadian Twitter users disclose health experiences related to influenza and COVID-19 — such as potential exposure, symptoms, treatment and disease outcomes — that could be indicative of recent plausible influenza cases, ILIs or symptomatic COVID-19 infections.…”

Section: Introductionmentioning

confidence: 99%

Comparison of pretrained transformer-based models for influenza and COVID-19 detection using social media text data in Saskatchewan, Canada

Tian

Zhang

Duan

et al. 2023

Front. Digit. Health

View full text Add to dashboard Cite

BackgroundThe use of social media data provides an opportunity to complement traditional influenza and COVID-19 surveillance methods for the detection and control of outbreaks and informing public health interventions.ObjectiveThe first aim of this study is to investigate the degree to which Twitter users disclose health experiences related to influenza and COVID-19 that could be indicative of recent plausible influenza cases or symptomatic COVID-19 infections. Second, we seek to use the Twitter datasets to train and evaluate the classification performance of Bidirectional Encoder Representations from Transformers (BERT) and variant language models in the context of influenza and COVID-19 infection detection.MethodsWe constructed two Twitter datasets using a keyword-based filtering approach on English-language tweets collected from December 2016 to December 2022 in Saskatchewan, Canada. The influenza-related dataset comprised tweets filtered with influenza-related keywords from December 13, 2016, to March 17, 2018, while the COVID-19 dataset comprised tweets filtered with COVID-19 symptom-related keywords from January 1, 2020, to June 22, 2021. The Twitter datasets were cleaned, and each tweet was annotated by at least two annotators as to whether it suggested recent plausible influenza cases or symptomatic COVID-19 cases. We then assessed the classification performance of pre-trained transformer-based language models, including BERT-base, BERT-large, RoBERTa-base, RoBERT-large, BERTweet-base, BERTweet-covid-base, BERTweet-large, and COVID-Twitter-BERT (CT-BERT) models, on each dataset. To address the notable class imbalance, we experimented with both oversampling and undersampling methods.ResultsThe influenza dataset had 1129 out of 6444 (17.5%) tweets annotated as suggesting recent plausible influenza cases. The COVID-19 dataset had 924 out of 11939 (7.7%) tweets annotated as inferring recent plausible COVID-19 cases. When compared against other language models on the COVID-19 dataset, CT-BERT performed the best, supporting the highest scores for recall (94.8%), F1(94.4%), and accuracy (94.6%). For the influenza dataset, BERTweet models exhibited better performance. Our results also showed that applying data balancing techniques such as oversampling or undersampling method did not lead to improved model performance.ConclusionsUtilizing domain-specific language models for monitoring users’ health experiences related to influenza and COVID-19 on social media shows improved classification performance and has the potential to supplement real-time disease surveillance.

show abstract

Vivid London: Assessing the resilience of urban vibrancy during the COVID-19 pandemic using social media data

Chen,

Liu,

et al. 2024

Sustainable Cities and Society

View full text Add to dashboard Cite

A chronological and geographical analysis of personal reports of COVID-19 on Twitter from the UK

Cited by 7 publications

References 25 publications

Social media engagement in health and climate change: an exploratory analysis of Twitter

Social media engagement in health and climate change: an exploratory analysis of Twitter

Comparison of pretrained transformer-based models for influenza and COVID-19 detection using social media text data in Saskatchewan, Canada

Vivid London: Assessing the resilience of urban vibrancy during the COVID-19 pandemic using social media data

Contact Info

Product

Resources

About