Banca de DEFESA: Pedro Henrique Borges da Silva

Natual language Processing, speech, complex networks.

The objective of the proposed work is to study the similarity between the networks of senators formed based on their voting patterns and that formed based on their speeches. To achieve this, two networks need to be constructed. For the voting network, connections are established between senators, with stronger connections between those who vote together more frequently. The speech network utilizes natural language processing methods and a similarity calculation technique based on textual data. A distance calculation method is then used to compare the two networks. By conducting this analysis for various legislatures, it will be possible to evaluate the consistency between politicians' speeches and voting patterns, as well as how this relationship has evolved over time.