Enhance Robustness of Language Models Against Variation Attack through Graph Integration
This addresses the problem of adversarial attacks in Chinese NLP, which is incremental as it builds on existing robust language model research with a domain-specific focus.
The study tackled the vulnerability of pre-trained language models to character variation attacks in Chinese content by proposing CHANGE, a method that integrates a Chinese character variation graph, resulting in improved robustness as shown in experiments across multiple NLP tasks.
The widespread use of pre-trained language models (PLMs) in natural language processing (NLP) has greatly improved performance outcomes. However, these models' vulnerability to adversarial attacks (e.g., camouflaged hints from drug dealers), particularly in the Chinese language with its rich character diversity/variation and complex structures, hatches vital apprehension. In this study, we propose a novel method, CHinese vAriatioN Graph Enhancement (CHANGE), to increase the robustness of PLMs against character variation attacks in Chinese content. CHANGE presents a novel approach for incorporating a Chinese character variation graph into the PLMs. Through designing different supplementary tasks utilizing the graph structure, CHANGE essentially enhances PLMs' interpretation of adversarially manipulated text. Experiments conducted in a multitude of NLP tasks show that CHANGE outperforms current language models in combating against adversarial attacks and serves as a valuable contribution to robust language model research. These findings contribute to the groundwork on robust language models and highlight the substantial potential of graph-guided pre-training strategies for real-world applications.