Annual Computer Security Applications Conference (ACSAC) 2021

Full Program »

Eluding ML-based Adblockers With Actionable Adversarial Examples

Online advertisers have been quite successful in circumventing traditional adblockers that rely on manually curated rules to detect ads. As a result, adblockers have started to use machine learning (ML) classifiers for more robust detection and blocking of ads. Among these, AdGraph which leverages rich contextual information to classify ads, is arguably, the state of the art ML-based adblocker. In this paper, we present A4, a tool that intelligently crafts adversarial ads to evade AdGraph. Unlike traditional adversarial examples in the computer vision domain that are somewhat unconstrained, adversarial ads generated by A4 are actionable in the sense that they preserve application semantics of the web page. Through a series of experiments we show that A4 can bypass AdGraph about 81% of the time, which surpasses the state-of-the-art attack by a significant margin of 145.5%, with an overhead of <20% and perturbations that are visually imperceptible in the rendered webpage. We envision that A4’s framework can be used to potentially launch adversarial attacks against other ML-based web applications.

Shitong Zhu
University of California, Riverside

Zhongjie Wang
University of California, Riverside

Xun Chen
Samsung Research America

Shasha Li
University of California, Riverside

Keyu Man
University of California, Riverside

Umar Iqbal
University of Iowa

Zhiyun Qian
University of California, Riverside

Kevin Chan
US Army Research Laboratory

Srikanth Krishnamurthy
University of California, Riverside

Zubair Shafiq
University of California, Davis

Paper (ACM DL)

Slides

Video

 



Powered by OpenConf®
Copyright©2002-2021 Zakon Group LLC