Abstract
"本文介绍了我们在第二十四届中文计算语言学大会细粒度中文仇恨言论识别任务中的参赛系统。该任务要求构建结构化仇恨四元组(评论对象、论点、目标群体、是否仇恨),提升模型的细粒度检测与可解释性。我们基于大语言模型,首先评估了LoRA参数高效微调效果,优化了超参数配置;其次对标注数据进行结构化处理,增强数据规范性;最后优化提示词设计,引导模型生成准确的结构化输出。实验表明,三阶段优化提升了模型性能。"
- Anthology ID:
- 2025.ccl-2.49
- Volume:
- Proceedings of the 24th China National Conference on Computational Linguistics (CCL 2025)
- Month:
- August
- Year:
- 2025
- Address:
- Jinan, China
- Editors:
- Hongfei Lin, Bin Li, Hongye Tan
- Venue:
- CCL
- SIG:
- Publisher:
- Chinese Information Processing Society of China
- Note:
- Pages:
- 411–418
- Language:
- URL:
- https://aclanthology.org/2025.ccl-2.49/
- DOI:
- Bibkey:
- Cite (ACL):
- Fanjun Lin, Yanwei Zhang, Huang Yang, and Zhiyuan Yao. 2025. CCL25-Eval任务10系统报告:面向细粒度中文仇恨言论识别的大语言模型增强. In Proceedings of the 24th China National Conference on Computational Linguistics (CCL 2025), pages 411–418, Jinan, China. Chinese Information Processing Society of China.
- Cite (Informal):
- CCL25-Eval任务10系统报告:面向细粒度中文仇恨言论识别的大语言模型增强 (Lin et al., CCL 2025)
- Copy Citation:
- PDF:
- https://aclanthology.org/2025.ccl-2.49.pdf