An improved method for predicting evolutionary link in email network

The prediction of the evolutionary link in the email network is an important research direction in the field of network security. The weighted correlated Bayesian classification model is an extension of the Naive Bayesian classification model. In this paper, email network users were grouped by the characteristics of email content and the evolutionary links were sorted into two types: the link in the same issue group and between two issue groups respectively. By defining classification attributes for each type of evolutionary link and depending on the weighted correlated Bayesian classification model, an improved method for predicting evolutionary link was proposed. The result of experiment in email dataset showed that the accuracy and precision of the improved method is higher than Common Neighbor algorithm and Adamic-Adar algorithm.