Visual Knowledge Discovery and Criminal Justice. Insights from a Computational Crime Analysis Research

Extracting actionable knowledge from heterogeneous information enclosed in criminal proceedings is one of the main challenges in the fast-growing field of computational crime analysis. Public prosecutors, on the other hand, can usually rely only on databases (containing complaints, criminal records, or police reports) accessible via traditional textual interfaces that lack advanced and visually-intuitive information extraction functionalities. This work presents the results of an ongoing research project that turns to visual knowledge discovery to help public prosecutors analyze structural and functional features of criminal networks under investigation. In this vein, the paper dwells on an experimental platform for computational crime analysis where visualization has three different goals: (a) highlighting the structural and qualitative features of both criminal organizations and their components; (b) depicting the progress and transformation of criminal networks over time; (c) enhancing the interaction between the legal scholar/expert and computational heuristics in the knowledge formation process. The outline of the solutions devised within the project—which has seen the involvement of domain experts and the use of real data—becomes an occasion to sketch a few insights about the potential role of visualization in criminal justice and, more in general, in the analysis of social phenomena