EigenCrime: An Algorithm for Criminal Network Mining Based on Trusted Computing

The searches on criminal network have become a key issue recent years, as the criminal activities are becoming grouping and networking. Most researches mainly focus on network of criminals, few on network of suspects and the traditional method is studying the static structure of the network, without explaining the behaviours among the nodes. In this paper, we are motivated by the ideal of EigenTrust algorithm to calculate the suspicion score of every member in the network of suspects and figure out the activists in the network by the score, supporting decision making for the police. Experiment results show high rank of criminals and low rank of innocents in a practical suspect network and prove the efficiency and accuracy of EigenCrime.