Homograph Attacks on Maghreb Sentiment Analyzers

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Qachfar, Fatima Zahra, Verma, Rakesh M.
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916115750322176
author Qachfar, Fatima Zahra
Verma, Rakesh M.
author_facet Qachfar, Fatima Zahra
Verma, Rakesh M.
contents We examine the impact of homograph attacks on the Sentiment Analysis (SA) task of different Arabic dialects from the Maghreb North-African countries. Homograph attacks result in a 65.3% decrease in transformer classification from an F1-score of 0.95 to 0.33 when data is written in "Arabizi". The goal of this study is to highlight LLMs weaknesses' and to prioritize ethical and responsible Machine Learning.
format Preprint
id arxiv_https___arxiv_org_abs_2402_03171
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Homograph Attacks on Maghreb Sentiment Analyzers
Qachfar, Fatima Zahra
Verma, Rakesh M.
Computation and Language
Cryptography and Security
Machine Learning
We examine the impact of homograph attacks on the Sentiment Analysis (SA) task of different Arabic dialects from the Maghreb North-African countries. Homograph attacks result in a 65.3% decrease in transformer classification from an F1-score of 0.95 to 0.33 when data is written in "Arabizi". The goal of this study is to highlight LLMs weaknesses' and to prioritize ethical and responsible Machine Learning.
title Homograph Attacks on Maghreb Sentiment Analyzers
topic Computation and Language
Cryptography and Security
Machine Learning
url https://arxiv.org/abs/2402.03171