HateVersarial: Adversarial Attack Against Hate Speech Detection Algorithms on Twitter

Edita Grolman, Hodaya Binyamini, Asaf Shabtai, Yuval Elovici, Ikuya Morikawa, Toshiya Shimizu

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

6 Scopus citations

Abstract

Machine learning (ML) models are commonly used to detect hate speech, which is considered one of the main challenges of online social networks. However, ML models have been shown to be vulnerable to well-crafted input samples referred to as adversarial examples. In this paper, we present an adversarial attack against hate speech detection models and explore the attack's ability to: (1) prevent the detection of a hateful user, which should result in termination of the user's account, and (2) classify normal users as hateful, which may lead to the termination of a legitimate user's account. The attack is targeted at ML models that are trained on tabular, heterogeneous datasets (such as the datasets used for hate speech detection) and attempts to determine the minimal number of the most influential mutable features that should be altered in order to create a successful adversarial example. To demonstrate and evaluate the attack, we used the open and publicly available "Hateful Users on Twitter" dataset. We show that under a black-box assumption (i.e., the attacker does not have any knowledge on the attacked model), the attack has a 75% success rate, whereas under a white-box assumption (i.e., the attacker has full knowledge on the attacked model), the attack has an 88% success rate.

Original languageEnglish
Title of host publicationUMAP2022 - Proceedings of the 30th ACM Conference on User Modeling, Adaptation and Personalization
PublisherAssociation for Computing Machinery, Inc
Pages143-152
Number of pages10
ISBN (Electronic)9781450392075
DOIs
StatePublished - 4 Jul 2022
Event30th ACM Conference on User Modeling, Adaptation and Personalization, UMAP2022 - Virtual, Online, Spain
Duration: 4 Jul 20227 Jul 2022

Conference

Conference30th ACM Conference on User Modeling, Adaptation and Personalization, UMAP2022
Country/TerritorySpain
CityVirtual, Online
Period4/07/227/07/22

Keywords

  • adversarial attack
  • hate speech
  • social media
  • Twitter

ASJC Scopus subject areas

  • Software

Fingerprint

Dive into the research topics of 'HateVersarial: Adversarial Attack Against Hate Speech Detection Algorithms on Twitter'. Together they form a unique fingerprint.

Cite this