A mail classification method and system combining user attributes

Abstract

The invention discloses a mail classification method and system combined with user attributes, which belong to the technical field of data mining. First, a classifier is used to initialize the probability of new mail spam, and then the user preference attribute of the new mail is matched in the built user attribute network. , email subject attribute, user intimacy attribute and user behavior attribute, and then use the influence of user preference attribute, mail subject attribute, user intimacy and user behavior attribute on spam probability to update the spam probability of email. The present invention also provides a system for implementing the above method. The present invention improves the accuracy of the existing mail classification technology by classifying and determining the mail in combination with the user attributes.

Description

technical field [0001] The invention belongs to the technical field of data mining, and more specifically relates to a method and system for classifying emails combined with user attributes. Background technique [0002] The development of the Internet has made e-mail an indispensable and important tool and communication medium in modern people's work and life. In life, we exchange information and remind affairs through email; in work, it is often used as an important channel for meeting requests, file transfers, and affairs notifications. [0003] While e-mail is convenient for everyone, the appearance of spam has brought great troubles to people's life and work. The harm that spam brings to e-mail users mainly includes wasting people's time, wasting equipment and network resources, and hindering the sending and receiving of normal mail. Some spam hides phishing websites, which induce users to click, thereby stealing user accounts and passwords Personal information violat...

AI Technical Summary

Problems solved by technology

The harm that spam brings to e-mail users mainly includes wasting people's time, wasting equipment and network resources, and hindering the sending and receiving of normal mail. Some spam hides phishing websites, which induce users to click, thereby stealing user accounts and passwords Personal information violates user privacy information, and there are cases of defrauding users of money by stealing user bank account information

[0004] At present, there are mainly the following technologies for mail classification at home and abroad: black-and-white list method, which is difficult to maintain black-and-white list, and the scalability is also poor, so it is only suitable as an auxiliary means combined with other technologies; keyword matching method , this technology requires the keyword database to be updated frequently to cope with these changes, which brings too much maintenance cost and has great limitations; mass filtering method, this technology is helpless in the face of spam that can constantly change the sending mailbox, only It is suitable as an auxiliary means combined with other technologies; Naive Bayesian classification method, by training Naive Bayesian classifier to classify emails, this technology has certain accuracy, but because the total number of spam emails is increasing year by year, we need a A Mail Classification Algorithm with Higher Accuracy

Images