Observatorio de I+D+i UPM

Memorias de investigación
Communications at congresses:
Social network analysis for fraud detection using complicity functions
Research Areas
  • Artificial intelligence
In social networks the actors are people o groups of people, whose connections are based on any form of social interaction between them, such as economic transactions. Social network analysis has important implications in the fight against organized crime, business associations with fraudulent purposes or terrorism. We propose several complicity functions to measure the degree of complicity between the actors in a social network with a set of previously identified fraudster actors, with the aim of detecting the partners of an organized crime plot. We also propose a procedure for ring detection consisting on five steps. The first and second steps compute the complicity of each actor with each fraudster actor and the strength of attraction between fraudster actors, respectively. The third step projects the set of toxic actors using multidimensional scaling in a plane. In the fourth step the points of the projection are grouped according to the DBSCAN algorithm. Finally, the fifth step adds each non-fraudster actor that maximizes its complicity. The ring detection procedure is illustrated with a real example including 835 linked companies, of which eight are fraudulent.
9th International Conference of the ERCIM WG on Computational and Methodological Statistics
Sevilla, España
Start Date
End Date
From page
To page
  • Autor: Antonio Jimenez Martin (UPM)
  • Autor: Alfonso Mateos Caballero (UPM)
Research Group, Departaments and Institutes related
  • Creador: Grupo de Investigación: Grupo de análisis de decisiones y estadística
  • Departamento: Inteligencia Artificial
S2i 2020 Observatorio de investigación @ UPM con la colaboración del Consejo Social UPM
Cofinanciación del MINECO en el marco del Programa INNCIDE 2011 (OTR-2011-0236)
Cofinanciación del MINECO en el marco del Programa INNPACTO (IPT-020000-2010-22)