Multilayer Network Analysis: The Identification of Key Actors in a Sicilian Mafia Operation

TL;DR

This paper constructs and analyzes a multilayer criminal network from an anti-mafia operation in Sicily, using multilayer network analysis to identify key actors involved in organized crime.

Contribution

It introduces a multilayer network model for criminal organizations and applies three different importance measures to identify key players.

Findings

Multilayer network effectively models complex criminal relationships.

Shared nodes across layers highlight central figures.

Analysis aids law enforcement in targeting key suspects.

Abstract

Recently, Social Network Analysis studies have led to an improvement and to a generalization of existing tools to networks with multiple subsystems and layers of connectivity. These kind of networks are usually called multilayer networks. Multilayer networks in which each layer shares at least one node with some other layer in the network are called multiplex networks. Being a multiplex network does not require all nodes to exist on every layer. In this paper, we built a criminal multiplex network which concerns an anti-mafia operation called "Montagna" and it is based on the examination of a pre-trial detention order issued on March 14, 2007 by the judge for preliminary investigations of the Court of Messina (Sicily). "Montagna" focus on two Mafia families called "Mistretta" and "Batanesi" who infiltrated several economic activities including the public works in the north-eastern part of Sicily, through a cartel of entrepreneurs close to the Sicilian Mafia. Originally we derived two single-layer networks, the former capturing meetings between suspected individuals and the latter recording phone calls. But some networked systems can be better modeled by multilayer structures where the individual nodes develop relationships in multiple layers. For this reason we built a two-layer network from the single-layer ones. These two layers share 47 nodes. We followed three different approaches to measure the importance of nodes in multilayer networks using degree as descriptor. Our analysis can aid in the identification of key players in criminal networks.