Detecting and disrupting criminal networks A data driven approach

Criminals organized in networks generate an estimated €900 billion a year worldwide, which is obtained at the cost of numerous human lives, economic development, social stability and democratic peace. The root of this global problem is retraceable to local social settings (e.g. neighborhoods, schools, pubs) in which different generations of (potential) criminals from various backgrounds find mutual trust to converge into networks. As compared to legitimate networks, criminal networks deliberately operate under covert conditions and outside the boundaries of law. Detecting and disrupting them is therefore considered one of the biggest challenges for law enforcement agencies across the globe. The aim of this thesis is to contribute to a better understanding of this complex reality by introducing a data-driven approach to the empirical study of organized crime.
The various studies in this thesis show that criminal networks operate as complex adaptive systems. A typical feature of such systems is that every individual actor can operate autonomously and interact with others at the same time, resulting in highly unpredictable outcomes. This thesis contributes to unraveling this complexity by introducing new methods to integrate different levels of data, to infer networks, and to make relevant selections of data in support of a data-driven approach. Based on our findings we can conclude that criminal network structures cannot be presumed but emerge. Quantitative analyses of these emerging networks can drive qualitative interpretations and assessment in order to seek rather than assume structure. As such we can consider this to be a paradigm shift in law enforcement practice as well as in organized crime research.