AI-enabled real-time social-media monitoring of public sentiments on COVID-19 vaccination

The onslaught of the COVID-19 pandemic, the hunt for a drug, and the successful development and administration of a vaccine has impacted people beyond measure worldwide. The availability of technology at hand enables one to research and assess the people’s attitudes under these trying circumstances. The results are generally employed to estimate the people's mood and their inclination towards policies and decisions.

Figure 2: Averaged weekly Facebook sentiment trends for (a) UK and (b) US

Averaged weekly Facebook sentiment trends for (a) UK and (b) US. Image Credit:

Specifically, such studies of the public attitude also help understand and develop the baseline confidence in vaccines. Analysts track online public discourse on platforms such as Facebook and Twitter platforms. It can help categorize and inform the development of demographic-level engagement and tailored communications strategies to promote diversity and inclusion in vaccination campaigns.

An interdisciplinary team used an artificial intelligence (AI)-based approach to watch and analyze the social-media public sentiment towards COVID-19 vaccinations. The team evaluated the content of Twitter and Facebook on the COVID-19 vaccination and published the work in the medRxiv preprint server.*

The authors found an overall averaged positive, negative, and neutral sentiment in the UK to be 58%, 22%, and 17%, compared to 56%, 24%, and 18% in the US, respectively. They also find similar overall averaged negative sentiments, on both platforms: for the UK it is 22.50% and for the US, it is 24.10%.

In this study, the authors find that the nine-month-long study shows averaged positive public sentiment towards the COVID-19 vaccination. They find that the posts and tweets reflect that the UK and the US's social media opinions are evolving, with both complementary and contrasting insights gleaned from these two platforms.

Previously, the governments relied on surveys for the information. Accurate data on public opinion was difficult to obtain. The data usually suffers from small sample sizes, closed questions, and limited Spatio-temporal granularity. To overcome these issues, social media data is employed. It increases numbers and enables real-time analyses of public sentiments and attitudes with considerable Spatio-temporal granularity.

In particular, there is significant untapped potential in drawing on AI-enabled social-media analysis to inform public policy research.”

To sieve out information from social media, established AI techniques, such as machine learning (ML), deep learning (DL), and natural language processing (NLP) are employed. Because social media data is mostly unstructured, it is amenable to extract topics and sentiments from the social media posts using these techniques.

The analysis involves categorizing sentiments into subjective opinions from the text, audio, and video to obtain the polarities and states of mind towards target topics, themes, or different aspects of interest. In a complementary approach, ‘stance detection’ is used, where a stance-label (favorable, against, none) is assigned to a post. The post may not refer to or even mention an opinion on the specific predetermined target (which is under study). However, the post may reflect the sentiment of the person.

Half the world's population are active on social media, including over 70% of the UK and US population. The online activity has also significantly increased during the pandemic; for example, by >37% for Facebook.

This study uses the most popular and representative social media platforms - Twitter and Facebook - in the English language. The Facebook posts and tweets posted in the UK and the US from 1 March to 22 November 2020 are used. For this study, over 158 million tweets are hydrated and utilized for analysis.

The authors observed that the difference between the averaged positive and negative sentiment trends was more pronounced on Facebook than Twitter. Interestingly, Twitter seems to be negatively biased. These sentiments relate to public apprehensions and concerns around delays or pauses in vaccine trials, vaccine safety, corporations, and governments influencing vaccine availability and rights exclusivity for economic benefits.

The authors graphically illustrate Facebook and Twitter sentiments (in terms of positive, negative, and neutral). They observe and discuss the key events impacting positive, negative, and neutral sentiments, and map the temporal trends.

They also show the geo-spatial mapping of the sentiments to states in the US and the UK. The authors find that the sentiments towards COVID-19 vaccination were:

  1. Negative in the West and Midwest of the country, namely: Idaho, Kansas, New Hampshire, West Virginia, Alabama
  2. Positive in the East, namely, Maine, Colorado, Georgia, Hawaii
  3. Cornwall, Kent, East Sussex, Surrey, and Dorset all in England, and Aberdeenshire, Angus, and Stirlingshire all in Scotland
  4. Most negative sentiment in West Sussex, Somerset, North Yorkshire, and Durham, all in England.

In conclusion, this study shows the temporal variations and the geo-spatial mappings in public sentiments on COVID-19 vaccination in the UK and the US.

They identify the public optimism over the vaccine development, effectiveness, and trials, including concerns over the safety, the economic viability, and the corporation control in the process.

The authors discuss the limitations of their study and the ramifications of the events (true, rumors, or false information) during the study period.

Due to the high demand for timely and trustworthy information about 2019- nCoV WHO technical risk communication and social media teams have been working closely to track and respond to myths and rumors.”

WHO Situation Report-13

This is an important study that shows how AI-enabled analysis throws light on the people's attitudes on social media and evaluates in which direction it leans towards. This study helps address vaccine skeptics' concerns and develop required public trust in immunization to realize the goal of herd immunity, the authors write.

*Important Notice

medRxiv publishes preliminary scientific reports that are not peer-reviewed and, therefore, should not be regarded as conclusive, guide clinical practice/health-related behavior, or treated as established information.

Journal reference:
  • Artificial intelligence-enabled analysis of UK and US public attitudes on Facebook and Twitter towards COVID-19 vaccinations. Amir Hussain, Ahsen Tahir, Zain Hussain, Zakariya Sheikh, Mandar Gogate, Kia Dashtipour, Azhar Ali, Aziz SheikhmedRxiv 2020.12.08.20246231; doi:
Dr. Ramya Dwivedi

Written by

Dr. Ramya Dwivedi

Ramya has a Ph.D. in Biotechnology from the National Chemical Laboratories (CSIR-NCL), in Pune. Her work consisted of functionalizing nanoparticles with different molecules of biological interest, studying the reaction system and establishing useful applications.


Please use one of the following formats to cite this article in your essay, paper or report:

  • APA

    Dwivedi, Ramya. (2020, December 14). AI-enabled real-time social-media monitoring of public sentiments on COVID-19 vaccination. News-Medical. Retrieved on March 08, 2021 from

  • MLA

    Dwivedi, Ramya. "AI-enabled real-time social-media monitoring of public sentiments on COVID-19 vaccination". News-Medical. 08 March 2021. <>.

  • Chicago

    Dwivedi, Ramya. "AI-enabled real-time social-media monitoring of public sentiments on COVID-19 vaccination". News-Medical. (accessed March 08, 2021).

  • Harvard

    Dwivedi, Ramya. 2020. AI-enabled real-time social-media monitoring of public sentiments on COVID-19 vaccination. News-Medical, viewed 08 March 2021,


The opinions expressed here are the views of the writer and do not necessarily reflect the views and opinions of News Medical.
You might also like... ×
Moderna vaccine reduces viral spread by at least 60 percent, finds study