Methodological Innovation in the Identification and Characterization of Brazilian Favelas

Abstract

This paper discusses the importance of innovative methodologies for the classification and characterization of favelas in Brazil, addressing the limitations of traditional approaches that fail to capture the complexity and diversity of these territories. By incorporating technologies such as high-resolution spatial data, machine learning, and Explainable Artificial Intelligence (XAI), it becomes possible to conduct more accurate analyses of the internal dynamics of favelas, thus promoting more effective public policies. The study emphasizes the need for collaboration among researchers, public administrators, and civil society to ensure that the proposed solutions are responsive to local needs. Furthermore, it explores the challenges involved in implementing these technologies, including data collection and the transparency of analytical models. The paper concludes that these methodologies have the potential to transform urban planning and foster a more inclusive and sustainable urbanization process.