Want to replicate these results or work with this data to see what insights you can find?
The nations of the world are in a race to beat the Covid-19 pandemic. The human impact of the Corona virus continues to be felt as the global number of those who have succumbed to the virus is now a staggering 1,7 million. Economically, Corona is wreaking havoc as many countries lack the financial reserves and healthcare systems they need to fight the virus or sustain their economies.
Could the Covid-19 vaccine be the end to all of this? Several early vaccine breakthroughs have been announced from large pharmaceuticals from their Phase 3 study recently:
On the 9th Nov 2020, Pfizer & Biontech announced the success of a vaccine candidate with over 90% efficacy.
On the 11th Nov 2020, National Research Center for Epidemiology and Microbiology, Russian Federation, demonstrated 92% efficacy.
On the 16th Nov 2020, Moderna announced the vaccine candidate with more than 94.5% efficacy.
On the 8th Dec 2020, Lancet published the analysis of the safety and efficacy of the ChAdOx1 nCoV-19 vaccine, with 70.4% efficacy.
But, how does the world feel about this? SVM Analytics uses AI to analyze sentiment
Anthony Damian, CEO of SVM Analytics, a business intelligence and technology solutions provider for over 15 years, who spearheaded this analysis said, “We wanted to understand the sentiment of the people of the world around Covid-19 and a vaccine. Particularly what the emotional temperature of people was after the announcement of the success of the vaccine candidates. These findings are unbiased and transparent, and hopefully can be used as a guide forward.”
Before the findings are revealed, let’s see how the data was obtained, processed, and analyzed.
The data source
Twitter feed data from around the world that was posted after the announcement of the success of the vaccine candidates against Covid-19 was extruded and analyzed. The data was extracted for the seven-day period after the announcement, from 18th Nov till 24th Nov 2020, and tweets with the words “Covid” and “vaccine” present were selected.
Qualifying sentiment
Within the context of Covid and vaccine, three different sentiments were identified: Positive, Negative, and Neutral.
Positive. Positive tweets are identified as those posted by people who believe in the vaccination and are willing to take the vaccine when it is released in the market.
Negative. These tweets are posted by people who do not believe in the vaccine, nor will they take the vaccine.
Neutral. These tweets are less emotive, and contain information around Covid-19 and a vaccine.
As this study is to determine feeling or sentiment, the analysis will focus primarily on negative and positive data results.
Here is an example of the different tweets by sentiment:
The data analysis tool
A leading artificial intelligence (AI) tool, SEDGE, was used to process the vast ocean of Twitter data against these sentiment criteria to quickly, efficiently, and deeply understand just exactly what is being said, and felt, about a proposed Covid-19 vaccine.
A glimpse into the future
After the data was analyzed, SEDGE provided a prediction of how the world would continue to respond to similar announcements going forward, based on the patterns identified in the data.
The findings
Most of the world feels negative. But the media is positive
The data was carefully analyzed for neutral, positive and negative sentiments, and it was found that, when positive and negative sentiment were stacked against each other, negative sentiment was in the overwhelming majority.
The number of negative tweets were more than the number of positive tweets about Covid and the Vaccine. The positive news was mainly from news media and in fact, most of the positive sentiment tweets came from news and media profiles.
The bulk of the tweets came from English-speaking nations, namely the United Kingdom and the U.S. When the data was split by country, the United Kingdom, France, and India showed a largely negative sentiment, where the U.S. showed more positive than negative. Canada showed an equal split between negative and positive.
Source: SVM Analytics: Country-wise split of positive, negative and neutral tweets
Then, to understand how strongly people felt about the vaccine, a polarity analysis was created. The findings showed that those who are pro-vaccine feel slightly more strongly about it than those who are against.
But who's behind the tweets?
To better understand the validity of the sentiment analysis, SVM Analytics looked at the user profiles of those who had tweeted.
When this data was filtered through SEDGE with positive and negative user sentiments, it is clear to see what is most important to the people posting the tweets by the most common words used in their user descriptions.
Source: SEDGE created Word Clouds to graphically illustrate user descriptions.
Many of the positive sentiment profiles were news sites, or users who are in the health industry.
Those who displayed a negative sentiment on the other hand were primarily invested in knowing and understanding the truth.
SVM Analytics took validation one step further, and filtered the most common words used in the tweets. Ignoring the common words such as vaccine(s) and Covid, these word clouds show the most common words used in tweets, grouped by sentiment. For positive sentiment, words such as social, false, lockdowns, Moderna, and Pfizer appear most. For negative, words such as false, tests, proven, useless, and fears appear most.
What to expect - Predictive analytics tells us
Predictive analytics is a complex process where patterns are identified in vast amounts of data. Then, the mathematical likelihood of these patterns occurring again is calculated. It is not fortune-telling, and it is not guess-work, it is rather a way of finding out what is very likely to occur in the future so that we can prepare for it timeously.
Based on the training of the existing data, SEDGE’s gradient boosting algorithm was used to predict potential sentiment within an 85% accuracy. Criteria that has a higher degree of influence over sentiment, such as user favorites, number of followers, hour of the day, number of friends, etc . were factored in.
The prediction? A rising number of those with a negative sentiment around the world.
The reasons behind the negativity and the way forward
Based on the content of the tweets, the majority are concerned about the side effects of the vaccine, and how unsafe it may be. They do not trust the positive tweets that come from the media, big pharma, and government, as there is no clarity into the basis for their statements, nor who could be financing these messages.
For the world to feel positive about the Covid-19 vaccination, they need to feel that the vaccine is safe, with no negative side effects. Should the studies with respect to efficacy of the vaccine be published in the public domain, people will be allowed to read and understand them for themselves, and will be able to make their own decision about vaccines.
If the detailed studies are not published by the vaccine manufacturers, public anxiety will continue to rise, and a lot of negative sentiment and global resistance will seep in the digital domain.
SVM Analytics’ CEO, Anthony Damian, concludes, “I hope that this brief analysis will spark new analysis on the safety aspect of the vaccines and bring medical research, statisticians, the medical community and government organizations together to provide clarity to the people of the world.”
The information or views expressed in this article is authentic to the best of our knowledge, and as such, it is prone to errors and the absence of some key information, for more information please review our Terms of Use.
Want to replicate these results or work with this data to see what insights you can find?