COVID-19 VACCINE INFODEMIC: SENTIMENT ANALYSIS OF THE TWITTER CONTENT
COVID-19 VACCINE INFODEMIC: SENTIMENT ANALYSIS OF THE TWITTER CONTENT

Summary/Abstract: Most of the content found on the internet is now produced by the user himself. This democratization of media platforms and the unprecedented interest in social networks are at the origin of text analysis methods, such as sentiment analysis, whose aim is to understand what a person feels and to analyze the polarity (positive, negative or neutral) of sentiments from texts posted online. However, this freedom of content creation and lack of regulation also comes with its weaknesses. Apart from a relatively accurate content, there is also a large proliferation of false opinions as well as messages created by spambots. From the start of the pandemic, a lot of misleading information circulate on social media. The aim of this paper is to use artificial intelligence techniques, such as sentiment analysis, in order to provide an insight into the fake news phenomenon behind the COVID-19 vaccine infodemic. We use the Twitter API to extract two corpora of tweets concerning vaccination, in Romanian and in English, and use a mix of Natural Language Processing techniques and Deep Learning techniques, from lexicon-based approaches to supervised learning algorithms, to explore the tweet content and to detect the feeling polarity.

• Details
• Contents