Environmental scanning of cocaine trafficking in Brazil: Evidence from geospatial intelligence and natural language processing methods

Cocaine trafficking threatens countries' national security and is a major public health challenge. Cocaine is transported from producer countries to consumer markets using various routes, methods, and transportation means. These routes develop in the geographical environment, are carefully planned and are geo-strategic objects that respond to the opportunities that drug trafficking organisations (DTOs) find to reduce the risks of interdiction. In this sense, individual drug seizure data (IDS) become essential indicators for identifying trends and understanding trafficking flows associated with drug trafficking routes. However, due to the illicit nature of DTOs, the availability of these data is considerably limited, hindering the ability to analyse and identify trends. This study presents a methodology for collecting and processing data from open-source information reported by Brazil's federal government news website. Using geospatial intelligence and natural language processing methods, we created a dataset with 939 records and 44 variables related to cocaine seizures in Brazil in 2022. We applied geospatial analysis techniques from this dataset to identify trends and potential cocaine trafficking flows. The results were broadly consistent with existing literature on drug trafficking. They demonstrated the potential of open-source information for environmental scanning and knowledge generation through geographic information science. The approach proposed in our research provides tools that can be used to complement drug trafficking monitoring and formulate public policies to strengthen prevention and enforcement strategies.