Malicious webpage advertising detection method based on advertising network topology

Abstract

The invention discloses a malicious webpage advertising detection method based on advertising network topology. The method comprises the basic steps of extraction of all URL in websites to be detected, extraction of URL relevant to advertisings, reconstruction of advertising re-orientation links, annotation of node characteristics, extraction of 3-node short-path sections and detection rule generation based on machine learning. The malicious webpage advertising detection method does not need any modification to websites of publishers, is easy to implement, higher in applicability and does not depend on the contents of advertising webpages, so that the method has robustness to code obfuscation and shell adding. Compared with a single malicious webpage advertising detection technology, various types of malicious webpage advertising attacks such as phishing and click fraud can be detected by adopting the method, and the method has universality. For detection on malicious webpage advertisings utilizing advertising alliance, the method still has effectiveness.

Description

technical field [0001] The invention relates to the technical field of Internet security, in particular to a malicious webpage advertisement detection method based on the advertisement network topology. Background technique [0002] Today, the Internet has become the dominant channel for marketing and promotion, and we rarely fail to encounter advertisements when browsing the web; compared with traditional media, online advertisements are more convenient and economical, but online advertisements are increasingly being attacked by attackers For illegal purposes, such as distributing malware, posting false information, click fraud, etc. Online advertising mainly includes three roles: publishers, who display advertisements on their web pages on behalf of advertisers; advertisers, who create advertisements, and are the source of income for online advertising. In the process of publishing advertisements, advertising networks connect publishers and advertisers In general, large a...

AI Technical Summary

Problems solved by technology

However, publishers of malicious web advertisements often use obfuscation and code packing techniques to increase the difficulty of detecting malicious web advertisements

Complicating the situation is the current very common ad network, which is a business model in which an ad network sells and resells the space it acquires from publishers to place ads on other ad networks and advertisers, in a process that is outsourced layer by layer Significantly increased chances of serving malicious web ads in

Images