Researching The Community Mining Based On Email System

The computer tech development boosts rapid growth of the network info. As a daily communication tool, E-mail embodies a mass of information. How to extracted valuable information from large-scale mail effectively and rapidly, and analysis whom is whole show in the network community constituted by email user, has become current challenging research topic.This text discuss around the pre-processing of email system & email network community discovery, systematic study the method of email extraction, classification, reduction, and the discovery of community core & level after pre-processing. Put forward email system data pro-processing tech which based on email system time attribute classification and attribute value reduction method, and community core discovery method which based on node core influence degree level.E-mail system based on the time attribute classification and attribute value reduction method is to classified e-mail system by the time attribute, and make attribute reduction via value mini support of email core attribute, to simplify recording and reduce the redundant data. The community core discovery method based on node core influence degree is linked from the node coverage, influence level and differences with the other nodes, total measure all the network nodes, and discover the core node. Finally, applied the data pre-processing and community analysis method to Enron email data together, find the Enron's internal community core and level. Test result show the scientific and effective discover the Enron's important figure proposed in this paper, consistent with the existing facts.This text put forward method of pre-processing, discovery of community core & level. It can extract valuable information exactly and discover community core & level effectively. This method has important significance in the prevention of terrorist attacks, network security research and social actors in the relationship.