Adversarial Style Augmentation via Large Language Model for Robust Fake News Detection
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2024
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866910913725988864 |
|---|---|
| author | Park, Sungwon Han, Sungwon Xie, Xing Lee, Jae-Gil Cha, Meeyoung |
| author_facet | Park, Sungwon Han, Sungwon Xie, Xing Lee, Jae-Gil Cha, Meeyoung |
| contents | The spread of fake news harms individuals and presents a critical social challenge that must be addressed. Although numerous algorithmic and insightful features have been developed to detect fake news, many of these features can be manipulated with style-conversion attacks, especially with the emergence of advanced language models, making it more difficult to differentiate from genuine news. This study proposes adversarial style augmentation, AdStyle, designed to train a fake news detector that remains robust against various style-conversion attacks. The primary mechanism involves the strategic use of LLMs to automatically generate a diverse and coherent array of style-conversion attack prompts, enhancing the generation of particularly challenging prompts for the detector. Experiments indicate that our augmentation strategy significantly improves robustness and detection performance when evaluated on fake news benchmark datasets. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2406_11260 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Adversarial Style Augmentation via Large Language Model for Robust Fake News Detection Park, Sungwon Han, Sungwon Xie, Xing Lee, Jae-Gil Cha, Meeyoung Computation and Language Artificial Intelligence The spread of fake news harms individuals and presents a critical social challenge that must be addressed. Although numerous algorithmic and insightful features have been developed to detect fake news, many of these features can be manipulated with style-conversion attacks, especially with the emergence of advanced language models, making it more difficult to differentiate from genuine news. This study proposes adversarial style augmentation, AdStyle, designed to train a fake news detector that remains robust against various style-conversion attacks. The primary mechanism involves the strategic use of LLMs to automatically generate a diverse and coherent array of style-conversion attack prompts, enhancing the generation of particularly challenging prompts for the detector. Experiments indicate that our augmentation strategy significantly improves robustness and detection performance when evaluated on fake news benchmark datasets. |
| title | Adversarial Style Augmentation via Large Language Model for Robust Fake News Detection |
| topic | Computation and Language Artificial Intelligence |
| url | https://arxiv.org/abs/2406.11260 |