PEREIRA, I. S.; http://lattes.cnpq.br/4984574616776730; PEREIRA, Igor de Sousa.
Abstract:
The Internet has enabled greater interaction between people around the world, and Twitter has
played a significant role in this. As one of the most widely used social networks on the planet, Twitter
allows users to make posts expressing various aspects of their daily lives. With the approach of major
events of worldwide recognition (e.g. sporting events), it is natural that users also express their
comments on the various aspects of these competitions. However, due to the huge amount of
comments generated daily, it is possible to observe a variety of interpretations that reflect different
aspects of the event, such as discussions about coaches' choices, teams' performance, and even
personal comments about the participants. This paper presents an approach to analyze the
discussions surrounding an event, using the World Cup as an example to validate the method. For this
purpose, data collected from Twitter was employed, which was used as input in clustering techniques
in order to identify potential sets of comments related to the competition. As a result, 13 diverse
clusters were obtained, each with unique characteristics, ranging from individual assessments, to
criticism of the host country Qatar, to team rivalries, among other aspects. These results indicate the
existence of patterns in comments on a specific theme, suggesting that users seek to comment on
topics of the moment and with high engagement.