Complicity Functions for Detecting Organized Crime Rings
Graph theory is an evident paradigm for analyzing social networks, which are the main tool for collective behavior research, addressing the interrelations between members of a more or less well-defined community. Particularly, social network analysis has important implications in the fight against organized crime, business associations with fraudulent purposes or terrorism.
Classic centrality functions for graphs are able to identify the key players of a network or their intermediaries. However, these functions provide little information in large and heterogeneous graphs. Often the most central elements of the network (usually too many) are not related to a collective of actors of interest, such as be a group of drug traffickers or fraudsters. Instead, its high centrality is due to the good relations of these central elements with other honorable actors.
In this paper we introduce complicity functions, which are capable of identifying the intermediaries in a group of actors, avoiding core elements that have nothing to do with this group. These functions can classify a group of criminals according to the strength of their relationships with other actors to facilitate the detection of organized crime rings.
The proposed approach is illustrated by a real example provided by the Spanish Tax Agency, including a network of 835 companies, of which eight were fraudulent.
KeywordsMultidimensional Scaling Social Network Analysis Geodesic Distance Business Association Complicity Function
The paper was supported by the project MTM2014-56949-C3-2-R.
- 1.Bavelas, A., Barret, D.: An experimental approach to organizational communication. Personnel 27, 366–371 (1951)Google Scholar
- 2.Cohn, B.S., Marriott, M.: Networks and centers in the integration in Indian civilization. J. Soc. Res. I(1), 1–9 (1958)Google Scholar
- 3.Czepiel, J.A.: Word of mouth processes in the diffusion of a major technological innovation. J. Mark. Res. I, 1–9 (1974)Google Scholar
- 5.Ester, M., Kriegel, H.-P., Sander, J., Xu, X.: A density-based algorithm for discovering clusters in large spatial databases with noise. In: KDD 1996 Proceedings (1996)Google Scholar
- 9.Hu, Y.: Efficient and high quality force-directed graph. Math. J. 10(1), 37–71 (2005)Google Scholar
- 15.Lawrence, P., Sergey, B., Rajeev, M., Terry, W.: The pagerank citation ranking: bringing order to the web. Technical report, Stanford University (1998)Google Scholar
- 16.Leavit, H.J.: Some effects of certain communication pattern on group performance. Massachusetts Institute of Technology (1949)Google Scholar
- 17.Peixoto, T.P.: Hierarchical block structures and high-resolution model selection in large networks. Phys. Rev. X 4(1), 011047 (2014)Google Scholar
- 21.Smith, S.L.: Communication Pattern and the Adaptability of Task-Oriented Groups: An Experimental Study. Group Networks Laboratory, Research Laboratory of Electronics, Massachusetts Institute of Technology, Cambridge (1950)Google Scholar