Methodology
The basis of DIXIM's work was the development of the Multidimensional Crime Analysis Method, the AMC Method. The method was structured in the analysis of qualitative data and victims' reports in thousands of events, over 15 years of research, with the objective of identifying the criminal problem that gave rise to the event.
Every record or news of a crime already has a police dimension (D1) and a legal dimension (D2), as provided for in the legislation. Useful information for criminal proceedings, but not for building a public policy or loss and risk management capabilities of controlling the criminal problem that gave rise to the crime. For this reason, we structured three other dimensions, which reveal: the economic objective (D5), demonstrating illicit markets and chains; the situation (D4) that facilitated the occurrence of the event; and the environment (D3), which attracted or encouraged the event, whether physical (environmental dimension) or virtual.
In order to operationalize the AMC Method, we have developed, since 2014, information technology solutions in the field of cognitive computing, capable of capturing and merging criminal events from different sources. They are natural language processing (NLP) and machine learning (ML) algorithms created to interpret events and regroup them into categories of criminal problems, that is, in their economic, situational, and environmental dimensions.