Detection of user relationships in networks using self-organization approach

For the purpose of monitoring and knowing the extent of use of objects in the network, generally each network operation is stored for subsequent analysis to determine which objects and how often are they accessed and modified. The self-organization of such multi-state operations is used in this work to help clustering of relationships amongst users or between users and network objects; and enhance the performance of network management. The attributes of involved users used in clustering helps in generating alerts well in advance due to misuse of objects. The experimental setup is built to demonstrate the effectiveness of this approach.