Studying social media can help understand public perception on various subjects, such
as healthcare, and provide valuable information for the time of the respective study.
Nowadays, the global pandemic of COVID-19 has resurfaced the subject of vaccination
and the contradiction of people against it. Insights from the analysis of social media
networks can help researchers understand the extend of vaccination awareness among
the public. A way to do so, is by using sentiment analysis. This process can be implemented by using TextBlob and VADER, which are two libraries in Python that can
evaluate a given text and return its sentiment score for each tweet. This score ranges
from -1,0 to +1,0, with -1,0 being extremely negative, +1,0 being extremely positive
and tweets with a sentiment score ranging from -0,1 to +0,1 being neutral. This study
uses 994.716 Twitter posts from eight hashtags about the top four COVID-19 vaccine
production companies and two hashtags about the antivaccination movement. The time
frame of the collection of tweets is from the 15th of July 2021 to the 7th of November
2021 (a total of 116 days). The aim is to analyze and compare the results of the two libraries, when applied on the same datasets. Results indicate that there were 92.064 special cases where the label was positive on one lexicon and negative on the other. Then,
by arbitrarily selecting ten tweets (one for each hashtag) from these special cases, custom scores are suggested with the aim of understanding the function of these tools, and
based on that, propose some potential improvements.
Collections
Show Collections
