Análise exploratória dos dados da Central de Atendimento à Mulher

Abstract

In recent years, there has been a worrying increase in violence against women. Faced with this issue, this work aims to collect, process, and analyze data from the Central de Atendimento à Mulher (Ligue 180) in order to create a dashboard containing key information about the victims, suspects, and reports. To achieve this, ETL techniques (Extraction, Transformation, and Loading of Data) were used. Python language was employed for data extraction, with the data being stored in BigQuery and processed using SQL. Finally, Power BI was used to construct data visualizations. Based on the insights and information obtained, it is noted that there is a recurring issue in the recording of complaint information, with many entries filled as ’Not Informed.’ This hinders the creation of complete profiles about the most affected groups. To mitigate this issue, it would be important to implement measures to ensure more complete and accurate data collection, such as training for those who record complaints and reviewing forms to include mandatory fields, so that specific measures can be directed to assist the most vulnerable populations. Despite this problem, it was possible to generate relevant information. The research revealed that of the reported data, 78% of complaints occurred in the victim’s own home. Additionally, it was observed that 91% of victims use telephone support channels, highlighting the importance of maintaining and expanding the "Ligue 180"service. The insights provided by the dashboards can be used by public policies that work in the defense of women to direct resources and efforts more effectively. For example, data on the high frequency of aggressions in domestic environments can help formulate awareness campaigns focused on family settings.