Swarm‏ Intelligent Agents of E-mail Classification for Contribution of Cyber and Forensic System

Abstract

Due to the Internet is an open, direct transmission tool and available to all for transmitting and exchanging messages, including those circulating among criminals, terrorists and those with improper motives towards others, therefore it became necessary to classify these messages by intelligent agent as a forensic evidence. In light of this, this paper included the urgent need to find a self-classified analysis of e-mail and to clarify what is suspicious from it, then classify the fraud Issuers which is covered in this paper. Sample data is a set of real e-mails. Some virtual messages were added to enhance the accuracy of the results obtained. The proposed work relies on swarm intelligent agents and modification of Voronoi algorithm such that the issues of the messages, including suspicious messages, are divided into communities. Moreover, Communities are divided into categories, each given a specific rank, depending on the quality and size of the threat messages. Moreover, this technique can be applied to messages and comments from people who talk about a particular subject and participants in the social media pages.

Keywords: Forensic; Voronoi; Classifier; SPF, Entropy.