Key points are not available for this paper at this time.
This study investigates the dynamics of network Compression and Rarefaction in electoral contexts, focusing on the reduction or strengthening of internal cohesion and connectivity among groups and discourses within Telegram groups during Brazil’s 2024 municipal elections. We employ Social Network Analysis (SNA) and Topological Data Analysis (TDA) techniques on temporal series to identify interaction patterns and persistent homological structures. These approaches reveal both the strengthening of clusters and the weakening of connections, influenced by the impact of disinformative narratives. Additionally, Content Analysis supports the interpretation of the networks and topologies formed throughout the analyzed period. Preliminary results identify Nuclear Constellations, characterized by highly cohesive networks centered on political campaigns, and Bipolar Constellations, composed of isolated clusters reinforcing ideological bubbles. The methodology builds on previous studies applying TDA to political communication in temporal series (Rocha, 2024b; Rocha, Silva, Mielli, 2024), adapting and expanding its application to the Brazilian electoral scenario.
Rocha et al. (Tue,) studied this question.