Banca de DEFESA: ALEX MATIAS GOMES

Uma banca de DEFESA de MESTRADO foi cadastrada pelo programa.
STUDENT : ALEX MATIAS GOMES
DATE: 20/05/2022
TIME: 13:00
LOCAL: videochamada: https://meet.google.com/caf-kjuu-fkt
TITLE:

Sentiment Analysis of User-generated Content on Twitter from political communications about Covid-19 vaccine  

KEY WORDS:

sentiment analysis; political communication; user-generated content; Covid-19.

PAGES: 131
BIG AREA: Ciências Sociais Aplicadas
AREA: Administração
SUMMARY:

The new coronavirus pandemic changed the lives of millions of people and was an important milestone for the growth of social media and User-generated Content in these media. Among the various topics related to the pandemic that were highlighted on social media, the vaccine against Covid-19 was one of the most addressed topics. In Brazil, speeches and political positions about vaccines generated reactions on the networks, which could influence the population's behavior about vaccination, thus this study was carried out with the objective of analyzing the communications of Brazilian politicians and the User-generated Content about the vaccine against to Covid-19 on Twitter. Based on 3907 tweets from politicians and 146,536 tweets in response to these communications, collected through the Snscrape application, two analyzes were carried out, one of quantitative content and one of sentiments. The Content Analysis, divided by groups of politicians, showed that the most frequent concepts mentioned the logistics of vaccines, mandatory vaccination and criticism of the president, whereas the Sentiment Analysis, carried out through an unsupervised method using the Oplexicon lexicon in Portuguese showed that most of the contents had a negative sentiment with 35.5%, revealing disagreement with the opinions of political communications, while positive and neutral sentiment were respectively 34% and 30.5%, which showed an intensification of feelings. In a segmented analysis, it was observed that a group of politicians who defended topics considered anti-vaccination had proportionally more content with negative sentiment. The conclusion highlights the uniqueness of this study, the contribution to the research of a flowchart for choosing a Sentiment Analysis method and the indication of future researches. Keywords: sentiment analysis; political communication; user-generated content; Covid-19.

BANKING MEMBERS:
Externo à Instituição - GABRIEL RODRIGO GOMES PESSANHA - UNIFAL-MG
Presidente - 1372521 - MARIA VALERIA PEREIRA DE ARAUJO
Interno - 2859852 - MAX LEANDRO DE ARAUJO BRITO