Method and system for identifying potential malicious user in website

Abstract

The invention provides a method for identifying a potential malicious user in a website. The method comprises the following steps: based on user basic data and service data, generating a community network map by using a FastUnfolding algorithm; obtaining community information from the community network map, wherein the community information comprises a community attribute and a map attribute of the community network map; generating a wide table based on the user basic data, the service data and the community information of the belonging community, and generating a derivative field based on the wide table, wherein the derivative field is a field that can be used for judging whether the user is a malicious user; using the feature data selected for training of the user as training data, and modeling by using a random forest algorithm to obtain a random forest model, wherein the feature data comprise basic information of the user, the community information of the user and the corresponding derivative field; and identifying the malicious user by using the random forest model. The invention further provides an apparatus for identifying the potential malicious user in the website.

Description

technical field [0001] The present invention relates generally to data analysis, and more particularly to a method and system for identifying potentially malicious users of a website. Background technique [0002] With the popularization of network technology, more and more people begin to use the network for social activities, entertainment and shopping, which brings greater challenges to website management. In particular, malicious users, that is, users with abnormal operations (such as users who maliciously receive coupons, maliciously place orders, maliciously post operations, etc.), may threaten the safe and effective operation of the website. Therefore, identifying potential malicious users is of great significance to ensure the normal and smooth operation of the website. [0003] The current malicious user identification methods are generally based on basic index modeling, index integration based on basic data table information, and a wide table is generated; then ba...

AI Technical Summary

Problems solved by technology

The disadvantage of this method is that the statistical dimension is relatively single, and the information retained is relatively small, which is not conducive to identifying potential malicious users

Images