Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2411.08248 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866913576834301952 |
|---|---|
| author | Li, Jiyao Ni, Mingze Gong, Yongshun Liu, Wei |
| author_facet | Li, Jiyao Ni, Mingze Gong, Yongshun Liu, Wei |
| contents | Deep learning underpins most of the currently advanced natural language processing (NLP) tasks such as textual classification, neural machine translation (NMT), abstractive summarization and question-answering (QA). However, the robustness of the models, particularly QA models, against adversarial attacks is a critical concern that remains insufficiently explored. This paper introduces QA-Attack (Question Answering Attack), a novel word-level adversarial strategy that fools QA models. Our attention-based attack exploits the customized attention mechanism and deletion ranking strategy to identify and target specific words within contextual passages. It creates deceptive inputs by carefully choosing and substituting synonyms, preserving grammatical integrity while misleading the model to produce incorrect responses. Our approach demonstrates versatility across various question types, particularly when dealing with extensive long textual inputs. Extensive experiments on multiple benchmark datasets demonstrate that QA-Attack successfully deceives baseline QA models and surpasses existing adversarial techniques regarding success rate, semantics changes, BLEU score, fluency and grammar error rate. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2411_08248 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Deceiving Question-Answering Models: A Hybrid Word-Level Adversarial Approach Li, Jiyao Ni, Mingze Gong, Yongshun Liu, Wei Computation and Language Artificial Intelligence Machine Learning Deep learning underpins most of the currently advanced natural language processing (NLP) tasks such as textual classification, neural machine translation (NMT), abstractive summarization and question-answering (QA). However, the robustness of the models, particularly QA models, against adversarial attacks is a critical concern that remains insufficiently explored. This paper introduces QA-Attack (Question Answering Attack), a novel word-level adversarial strategy that fools QA models. Our attention-based attack exploits the customized attention mechanism and deletion ranking strategy to identify and target specific words within contextual passages. It creates deceptive inputs by carefully choosing and substituting synonyms, preserving grammatical integrity while misleading the model to produce incorrect responses. Our approach demonstrates versatility across various question types, particularly when dealing with extensive long textual inputs. Extensive experiments on multiple benchmark datasets demonstrate that QA-Attack successfully deceives baseline QA models and surpasses existing adversarial techniques regarding success rate, semantics changes, BLEU score, fluency and grammar error rate. |
| title | Deceiving Question-Answering Models: A Hybrid Word-Level Adversarial Approach |
| topic | Computation and Language Artificial Intelligence Machine Learning |
| url | https://arxiv.org/abs/2411.08248 |