Identificação e análise de bolhas sociais na Universidades Federal do Ceará - Campus Russas por meio de aprendizado de máquina não supervisionado

Abstract

Throughout history, it is common for people to gravitate towards those who share similar interests and characteristics, whether in everyday life or in academic settings. However, the formation of groups and subgroups based on academic affinities, hobbies, cultural backgrounds, and likemindedness can lead to the creation of social bubbles. These social bubbles may have negative consequences, such as segregation and limited interaction among students. This hinders the development of a diverse and inclusive university community, making it difficult to exchange knowledge and understand different perspectives. In this study, we propose the identification and analysis of social bubbles at the Federal University of Ceará - Campus Russas, using unsupervised machine learning techniques to detect groupings among students. The results are analyzed through visual representations of the identified clusters, enabling the identification of patterns of proximity, affinity, and social isolation. These findings provide a deeper understanding of the social structure at the Russas Campus, revealing closely-knit groups and areas with reduced interaction.