A Mail Classification Method Combining User Relationship and Bayesian Theory

Abstract

The invention discloses a mail classification method combining user relationship and Bayesian theory. By extracting the user relationship contained in the mail to construct a user relationship graph and combining with the improved naive Bayesian method, automatic classification of emails is realized and the improvement of The accuracy rate of the classification system reduces the misjudgment rate. The present invention proposes a credibility factor to estimate the credibility of the Naive Bayesian classifier classification results, and combines the Naive Bayesian method with the user relationship diagram, and uses the user relationship contained in the normal mail to construct the user relationship diagram, And generate a user white list according to the user's general habit of handling emails. In the process of classifying new emails, the classification results are continuously fed back to the user relationship graph, and the user whitelist is updated at the same time, so that the classification system can automatically adjust the user relationship graph and whitelist according to changes in new emails, thereby achieving higher accuracy.

Description

technical field [0001] The invention belongs to the technical field of data mining, and more specifically relates to a mail classification method combining user relationship and Bayesian theory. Background technique [0002] With the rapid development of the Internet today, people's daily life has been integrated with the network environment. More and more people use the Internet for office, shopping, consumption, entertainment and other activities, among which email (E-mail) has become a daily One of the important means of communication. According to the 35th "Statistical Report on Internet Development in China" released by China Internet Network Information Center (CNNIC) in February 2015, as of December 2014, the number of Chinese netizens exceeded 649 million. More than 251 million. In foreign countries, there were about 929 million business email accounts in 2013, and it is still growing. However, problems also follow. A large number of spam mails are flooding people...

AI Technical Summary

Problems solved by technology

[0026] Aiming at the above defects or improvement needs of the prior art, the present invention provides a mail classification method combining user relationship and Bayesian theory. The technical problems of low classification accuracy and high misjudgment rate of normal emails, combined the user relationship contained in the emails with the Naive Bayesian method to improve the accuracy of mail classification and reduce the misjudgment rate of mails

Images