Identify the Leaders of Criminal Networks based on the Criminological System

Identify the Leaders of Criminal Networks based on the Criminological System

Zahra Gholamhoseini1 Sara Nazari2 Mehdi Fartash3

1) Department of Computer, Arak Branch, Islamic Azad University, Arak, Iran
2) Department of Computer, Arak Branch, Islamic Azad University, Arak, Iran
3) Department of Computer, Arak Branch, Islamic Azad University, Arak, Iran

IOI Code : XBGE-FDBDD
Publication : 11th International Conference on New Solutions in Engineering, Information Science and Technology of the Century ahead(icietconf.com)
Abstract :
    In this research, a new algorithm is proposed to identify leaders in criminal networks. The proposed algorithm includes three main steps: Determining the value of criminal network communications , Identifying the influential nodes in the criminal network and Updating the weight of network communications and determining the most common routes between leaders in the criminal network . In the proposed method, the process of identifying influence nodes (leaders) and the most commonly used paths between them is performed based on specific time periods. First, the criminal network is considered as a directional graph and the users of this network are described as network nodes. The first step in the proposed method is to determine the importance of communication between users by calculating the weight of the edges between network nodes. For this purpose, a criterion based on the information dissemination rate will be used. In the second step, the proposed method uses four criteria based on centrality to identify the importance of each user in the criminal network and determine the effective users in the network. This criterion ranks network nodes by simultaneously considering the data rate (weight of graph edges) and the number of active user communications. In the next step of the proposed method, the weight of network communications is modified using the data propagation rate and the impact rate of the nodes connected to the edge. After updating the weight of the network edges, the path identification operation is performed. The most frequently used paths between criminal network leaders can be considered as the most frequently used paths between influential network nodes, which is described using a minimal spanning tree structure. In order to evaluate the efficiency of the proposed method, its performance in identifying criminal network leaders has been investigated using the criteria of accuracy and processing time. Based on the results of the experiments, the proposed method over time can process the communications and messages exchanged between users and identify the affected nodes more accurately.

Keywords : social networking identifying leaders criminal networking social networking analysis