Result Analysis and Performance Evaluation of ACENET through Confusion Matrix Technique on Standard Dataset

The ACENET, first FIPA compliant agent framework over. NET to provide facility of abstract programming to agent oriented programmers, deployed over corporate network to monitor behavior of the users to generate and maintain profiles to avoid insider threat. Agents running over framework are intelligently performing their tasks in backend and become activate as users log into the system. According to local needs of the organizations the agents have been created to gather information of various activities inside organization. This paper presents brief overview of what and how developed framework monitors the behavior. The importance of organizationʹs policy has been discussed where weight-age has been assigned to profile attributes to identify threat contribution of each suspicious activity. The performance of the model has also been evaluated on the basis of specified parameters based on confusion matrix technique. Finally collected results have been analyzed to match with the targeted objectives