Towards a unified framework for imperceptible textual attacks

doi:10.1007/s10489-024-05292-6

CORC > 自动化研究所 > 中国科学院自动化研究所 > 多模态人工智能系统全国重点实验室

	Towards a unified framework for imperceptible textual attacks
	Shi, Jiahui; Li, Linjing; Zeng, Daniel
刊名	APPLIED INTELLIGENCE
	2024-02-09
页码	14
关键词	Adversarial attack Backdoor attack Natural language processing Adversarial machine learning
ISSN号	0924-669X
DOI	10.1007/s10489-024-05292-6
通讯作者	Li, Linjing(linjing.li@ia.ac.cn)
英文摘要	Despite the great success of Deep Neural Networks (DNNs) in the field of natural language processing (NLP), they are increasingly facing tremendous threats from textual attacks in two kinds: adversarial attacks and backdoor attacks. Both of them are able to manipulate DNNs into producing the designated target label. By searching the optimal replacement in the massive space of possible candidates, current textual attacks deal with each input sample one at a time. However, attacking in this manner is time consuming, and the generated samples suffer from low semantic consistency and language fluency. To address this issue, we design a unified framework for targeted adversarial attacks and backdoor attacks, which employs a masked language model to produce imperceptible poisoned samples directly. We conduct extensive experiments on three benchmark datasets for three different NLP model architectures. Experimental results reveal that the proposed framework can achieve the state-of-the-art attacking performance for backdoor attacks with a substantial improvement, and a more pronounced improvements for targeted adversarial attacks, while concurrently maintaining the high linguistic quality of generated samples.
资助项目	National Natural Science Foundation of China[XDA27030100] ; Strategic Priority Research Program of Chinese Academy of Sciences[72293573] ; Strategic Priority Research Program of Chinese Academy of Sciences[72293575] ; National Natural Science Foundation of China
WOS研究方向	Computer Science
语种	英语
出版者	SPRINGER
WOS记录号	WOS:001157397800001
资助机构	National Natural Science Foundation of China ; Strategic Priority Research Program of Chinese Academy of Sciences ; National Natural Science Foundation of China
内容类型	期刊论文
源URL	[http://ir.ia.ac.cn/handle/173211/55574]
专题	多模态人工智能系统全国重点实验室
通讯作者	Li, Linjing
作者单位	Chinese Acad Sci, Inst Automat, State Key Lab Multimodal Artificial Intelligence S, Beijing 100190, Peoples R China
推荐引用方式 GB/T 7714	Shi, Jiahui,Li, Linjing,Zeng, Daniel. Towards a unified framework for imperceptible textual attacks[J]. APPLIED INTELLIGENCE,2024:14.
APA	Shi, Jiahui,Li, Linjing,&Zeng, Daniel.(2024).Towards a unified framework for imperceptible textual attacks.APPLIED INTELLIGENCE,14.
MLA	Shi, Jiahui,et al."Towards a unified framework for imperceptible textual attacks".APPLIED INTELLIGENCE (2024):14.