Backdoors against natural language processing: a review

Shaofeng Li; Tian Dong; Benjamin Zi Hao Zhao; Minhui Xue; Suguo Du; Haojin Zhu

doi:10.1109/MSEC.2022.3181001

Backdoors against natural language processing: a review

Shaofeng Li, Tian Dong, Benjamin Zi Hao Zhao, Minhui Xue, Suguo Du, Haojin Zhu

School of Computing

Research output: Contribution to journal › Review article › peer-review

30 Citations (Scopus)

Abstract

Data poisoning attacks, specifically backdoor attacks, present a severe security threat in artificial intelligence. We provide a comprehensive survey into state-of-the-art backdoor attacks and defenses in the field of natural language processing.

Original language	English
Pages (from-to)	50-59
Number of pages	10
Journal	IEEE Security and Privacy
Volume	20
Issue number	5
DOIs	https://doi.org/10.1109/MSEC.2022.3181001
Publication status	Published - Sept 2022

Access to Document

10.1109/MSEC.2022.3181001

Cite this

@article{cfaa8b84cd7f4a3ea6f9154a0e991b61,

title = "Backdoors against natural language processing: a review",

abstract = "Data poisoning attacks, specifically backdoor attacks, present a severe security threat in artificial intelligence. We provide a comprehensive survey into state-of-the-art backdoor attacks and defenses in the field of natural language processing. ",

author = "Shaofeng Li and Tian Dong and Zhao, \{Benjamin Zi Hao\} and Minhui Xue and Suguo Du and Haojin Zhu",

year = "2022",

month = sep,

doi = "10.1109/MSEC.2022.3181001",

language = "English",

volume = "20",

pages = "50--59",

journal = "IEEE Security and Privacy",

issn = "1540-7993",

publisher = "Institute of Electrical and Electronics Engineers (IEEE)",

number = "5",

}

TY - JOUR

T1 - Backdoors against natural language processing

T2 - a review

AU - Li, Shaofeng

AU - Dong, Tian

AU - Zhao, Benjamin Zi Hao

AU - Xue, Minhui

AU - Du, Suguo

AU - Zhu, Haojin

PY - 2022/9

Y1 - 2022/9

N2 - Data poisoning attacks, specifically backdoor attacks, present a severe security threat in artificial intelligence. We provide a comprehensive survey into state-of-the-art backdoor attacks and defenses in the field of natural language processing.

AB - Data poisoning attacks, specifically backdoor attacks, present a severe security threat in artificial intelligence. We provide a comprehensive survey into state-of-the-art backdoor attacks and defenses in the field of natural language processing.

UR - https://www.scopus.com/pages/publications/85135756775

UR - http://purl.org/au-research/grants/arc/DP210102670

U2 - 10.1109/MSEC.2022.3181001

DO - 10.1109/MSEC.2022.3181001

M3 - Review article

AN - SCOPUS:85135756775

SN - 1540-7993

VL - 20

SP - 50

EP - 59

JO - IEEE Security and Privacy

JF - IEEE Security and Privacy

IS - 5

ER -

Backdoors against natural language processing: a review

Abstract

Access to Document

Other files and links

Fingerprint

Cite this