Generating Fluent Chinese Adversarial Examples for Sentiment Classification

被引:0
作者
Wang, Congyi [1 ,2 ]
Zeng, Jianping [1 ,2 ]
Wu, Chengrong [1 ,2 ]
机构
[1] Fudan Univ, Sch Comp Sci, Shanghai 200433, Peoples R China
[2] Minist Educ, Engn Res Ctr Cyber Secur Auditing & Monitoring, Shanghai 200433, Peoples R China
来源
2020 IEEE 14TH INTERNATIONAL CONFERENCE ON ANTI-COUNTERFEITING, SECURITY, AND IDENTIFICATION (ASID) | 2020年
基金
国家重点研发计划;
关键词
Adversarial examples; Chinese natural language; Sentiment classification;
D O I
暂无
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Highly accurate classifiers can be trained by existing machine learning models, however, most of these classifiers do not consider the adversarial attack. This makes these classifiers vulnerable to adversarial examples. In order to improve the ability of sentiment classifiers to resist the adversarial attack, it is very important to generate high-quality adversarial examples. Most of the existing methods that generate natural language adversarial examples aim at English text with relatively simple strategies, but a single transformation strategy is easily detected by the defender. In this paper, we propose a new method to generate Chinese natural language adversarial examples, which is called AD-ER (Adversarial Examples with Readability). The first step is to select the important words in the text, which have great impact on the sentiment classifier. Then we proposed four variant strategies to replace the important words and the best candidate word is selected heuristically under the constraints of its readability and maximum entropy model. The simulation results on a real shopping review dataset verify that the examples generated by our method can produce large attack disturbance to the classifiers. Different from other examples, our examples have good readability and diversity, which are more fluent and harder to be detected.
引用
收藏
页码:149 / +
页数:6
相关论文
共 50 条
[31]   An adversarial-example generation method for Chinese sentiment tendency classification based on audiovisual confusion and contextual association [J].
Ou, Hongxu ;
Yu, Long ;
Tian, Shengwei ;
Chen, Xin ;
Shi, Chen ;
Wang, Bo ;
Zhou, Tiejun .
KNOWLEDGE AND INFORMATION SYSTEMS, 2023, 65 (12) :5231-5258
[32]   Imbalanced sentiment classification based on sequence generative adversarial nets [J].
Wang, Chuantao ;
Yang, Xuexin ;
Ding, Linkai .
JOURNAL OF INTELLIGENT & FUZZY SYSTEMS, 2020, 39 (05) :7909-7919
[33]   SENTIMENT CLASSIFICATION ON CHINESE REVIEWS BASED ON AMBIGUOUS SENTIMENT CONFINED LIBRARY [J].
Liu, Meijuan ;
Yang, Shicai ;
Chen, Qiaofen .
2012 IEEE 2nd International Conference on Cloud Computing and Intelligent Systems (CCIS) Vols 1-3, 2012, :1470-1473
[34]   Exploiting effective features for chinese sentiment classification [J].
Zhai, Zhongwu ;
Xu, Hua ;
Kang, Bada ;
Jia, Peifa .
EXPERT SYSTEMS WITH APPLICATIONS, 2011, 38 (08) :9139-9146
[35]   Sentiment Classification Analysis of Chinese Microblog Network [J].
Wang, Xiaotian ;
Zhang, Chuang ;
Wu, Ming .
COMPLEX NETWORKS VI, 2015, 597 :123-129
[36]   A New Black Box Attack Generating Adversarial Examples Based on Reinforcement Learning [J].
Xiao, Wenli ;
Jiang, Hao ;
Xia, Song .
2020 INFORMATION COMMUNICATION TECHNOLOGIES CONFERENCE (ICTC), 2020, :141-146
[37]   Towards a Robust Classifier: An MDL-Based Method for Generating Adversarial Examples [J].
Asadi, Behzad ;
Varadharajan, Vijay .
2020 IEEE 19TH INTERNATIONAL CONFERENCE ON TRUST, SECURITY AND PRIVACY IN COMPUTING AND COMMUNICATIONS (TRUSTCOM 2020), 2020, :793-801
[38]   Targeted Adversarial Examples Generating Method Based on cVAE in Black Box Settings [J].
YU Tingyue ;
WANG Shen ;
ZHANG Chunrui ;
WANG Zhenbang ;
LI Yetian ;
YU Xiangzhan .
CHINESE JOURNAL OF ELECTRONICS, 2021, 30 (05) :866-875
[39]   Generating Adversarial Examples in One Shot With Image-to-Image Translation GAN [J].
Zhang, Weijia .
IEEE ACCESS, 2019, 7 :151103-151119
[40]   TextFirewall: Omni-Defending Against Adversarial Texts in Sentiment Classification [J].
Wang, Wenqi ;
Wang, Run ;
Ke, Jianpeng ;
Wang, Lina .
IEEE ACCESS, 2021, 9 :27467-27475