Geodesic
Translate your gibberish: black-box adversarial attack on machine translation systems
Zapiski Nauchnykh Seminarov POMI, Investigations on applied mathematics and informatics. Part II–2, Tome 530 (2023), pp. 96-112

Voir la notice de l'article provenant de la source Math-Net.Ru

Neural networks are deployed widely in natural language processing tasks on the industrial scale, and perhaps most often they are used as compounds of automatic machine translation systems. In this work, we present a simple approach to fool state of the art machine translation tools in the task of translation from Russian to English and vice versa. Using a novel black-box gradient-free tensor-based optimizer, we show that many online translation tools, such as Google, DeepL, and Yandex, may both produce wrong or offensive translations for nonsensical adversarial input queries and refuse to translate seemingly benign input phrases. This vulnerability may interfere with understanding a new language and simply worsen the user's experience while using machine translation systems, and, hence, additional improvements of these tools are required to establish better translation. 
@article{ZNSL_2023_530_a7,
     author = {A. Chertkov and O. Tsymboi and M. Pautov and I. Oseledets},
     title = {Translate your gibberish: black-box adversarial attack on machine translation systems},
     journal = {Zapiski Nauchnykh Seminarov POMI},
     pages = {96--112},
     publisher = {mathdoc},
     volume = {530},
     year = {2023},
     language = {en},
     url = {http://geodesic.mathdoc.fr/item/ZNSL_2023_530_a7/}
}
TY  - JOUR
AU  - A. Chertkov
AU  - O. Tsymboi
AU  - M. Pautov
AU  - I. Oseledets
TI  - Translate your gibberish: black-box adversarial attack on machine translation systems
JO  - Zapiski Nauchnykh Seminarov POMI
PY  - 2023
SP  - 96
EP  - 112
VL  - 530
PB  - mathdoc
UR  - http://geodesic.mathdoc.fr/item/ZNSL_2023_530_a7/
LA  - en
ID  - ZNSL_2023_530_a7
ER  - 
%0 Journal Article
%A A. Chertkov
%A O. Tsymboi
%A M. Pautov
%A I. Oseledets
%T Translate your gibberish: black-box adversarial attack on machine translation systems
%J Zapiski Nauchnykh Seminarov POMI
%D 2023
%P 96-112
%V 530
%I mathdoc
%U http://geodesic.mathdoc.fr/item/ZNSL_2023_530_a7/
%G en
%F ZNSL_2023_530_a7
A. Chertkov; O. Tsymboi; M. Pautov; I. Oseledets. Translate your gibberish: black-box adversarial attack on machine translation systems. Zapiski Nauchnykh Seminarov POMI, Investigations on applied mathematics and informatics. Part II–2, Tome 530 (2023), pp. 96-112. http://geodesic.mathdoc.fr/item/ZNSL_2023_530_a7/