Research And Development Of The E-mail Network Relationship Digging Method Based On Weighted And Directed Graph

Social Network Analysis is one of the hot issues among various study fields, relevant results are generated and put into practice. Email is a significant approach for people to communicate and share different opinions. As the network constituted by Email senders is in accordance with the features of the Complex Network, Email Network becomes one significant branch of the Social Network Analysis. In this dissertation, we would like to start the research on the association relationship among emails, discover the properties of community network, then deeper mining the real people behavior behind.The relationship between nodes in the Email Network is chosen as the study direction of the dissertation, main parts are as followed:1. Introduce the recent situation of Social Network Analysis(e.g. latest concepts, study result.etc.), and raise some representative algorithms.2. Analysis of mail network’s communication patterns, according to the email communication with direction and weight, construct the communication network to empower a message, to use for studying the genetic relationship.3. Based on the directed graph with weight, we have proposed a detection algorithm of local club boundary and mail information interaction. With The node important degree computing by the improved Pagerank algorithm, we can calculate the node which is closed with the target node.4. In the project, We have implemented the modules. According to the demand, it can push three noteworthy mail users. We tested the algorithm on the Random weighting network and the enron email dataset, Experiments show that this algorithm can accurately find the local community structure and the nodes closed to the target node in email networks. It is proved that the algorithm has a certain practicality.Through the analysis in Email network, we could understand the association relationship and specific role of each user. The work of this report provides some references for the further study of Email Entity Relationship...