Expanding Scope: Adapting English Adversarial Attacks to Chinese

TL;DR

This paper explores how to adapt English adversarial attack methods to Chinese NLP models, demonstrating their effectiveness in generating high-quality adversarial examples that improve model robustness.

Contribution

It introduces methods for adapting SOTA English adversarial attacks to Chinese, considering linguistic differences, and shows their effectiveness in generating fluent, semantically consistent adversarial examples.

Findings

English attack methods can be adapted to Chinese with proper segmentation and constraints

Adversarial examples in Chinese achieve high fluency and semantic consistency

Using these adversarial examples can enhance Chinese NLP model robustness

Abstract

Recent studies have revealed that NLP predictive models are vulnerable to adversarial attacks. Most existing studies focused on designing attacks to evaluate the robustness of NLP models in the English language alone. Literature has seen an increasing need for NLP solutions for other languages. We, therefore, ask one natural question: whether state-of-the-art (SOTA) attack methods generalize to other languages. This paper investigates how to adapt SOTA adversarial attack algorithms in English to the Chinese language. Our experiments show that attack methods previously applied to English NLP can generate high-quality adversarial examples in Chinese when combined with proper text segmentation and linguistic constraints. In addition, we demonstrate that the generated adversarial examples can achieve high fluency and semantic consistency by focusing on the Chinese language's morphology and phonology, which in turn can be used to improve the adversarial robustness of Chinese NLP models.