Application Research Of Data Mining Based On Social Networking Site

With the development of internet technology, email communication is one of the most frequentcommunications used by most users,so email communication network has become a kind ofnetwork resources can be a dircect response to people’s social behavior and globally presentssome orderly phenomena, for example, different hobbies or social relationships formed differentcommunities. Mining out the community structure of hunman society on the large amount of emailcommunication data and analyzing the importance of community members of society has becomean important research topic.This paper makes a study on analysis of email data, network construction, important discoveryand community detection and so on.the main work includes the following aspects:First， Email communication network attribute description. Analyzing the email data,Constructing Email network by extracting email communication relationship between the entities.Second，Important node discovery. Aiming at the features of email network connectivity, thispaper proposes an metrics index value of NR. Discovering important nodes by computing the NRvalue of node. Experiment on the Enron email data proves the way of computing NR value of nodesis superior to the way of finding important nodes based on graph theory.Third，Community detection based on edge weight on email network. Proposing a kind ofCommunity detection algorithm by seting important node as the core. Mining Community on emailnetwork,verifying the effect of Community Algorithm by doing the experiments in the Enron emaildata and Obtaining the community having close relationship with important node.Forth， Constructing relationship mining analysis prototype system based on emailcommunication network. Designing and implementing the module of important nod e discovery andcommunity detection.Finally this paper summarizes the work and presents further research aspects of anaylsis andmining on the email communication network.