Abstract
近年来,尽管深度学习给语义依存分析带来了长足的进步,但由于语义依存分析数据标注代价非常高昂,并且在单领域上性能较好的依存分析器迁移到其他领域时,其性能会大幅度下降。因此为了使其走向实用,就必须解决领域适应问题。本文提出一个新的基于对抗学习的领域适应依存分析模型,我们提出了基于对抗学习的共享双编码器结构,并引入领域私有辅助任务和正交约束,同时也探究了多种预训练模型在跨领域依存分析任务上的效果和性能。- Anthology ID:
- 2020.ccl-1.73
- Volume:
- Proceedings of the 19th Chinese National Conference on Computational Linguistics
- Month:
- October
- Year:
- 2020
- Address:
- Haikou, China
- Editors:
- Maosong Sun (孙茂松), Sujian Li (李素建), Yue Zhang (张岳), Yang Liu (刘洋)
- Venue:
- CCL
- SIG:
- Publisher:
- Chinese Information Processing Society of China
- Note:
- Pages:
- 783–794
- Language:
- Chinese
- URL:
- https://aclanthology.org/2020.ccl-1.73
- DOI:
- Cite (ACL):
- Dazhan Mao, Huayong Li, and Yanqiu Shao. 2020. 半监督跨领域语义依存分析技术研究(Semi-supervised Domain Adaptation for Semantic Dependency Parsing). In Proceedings of the 19th Chinese National Conference on Computational Linguistics, pages 783–794, Haikou, China. Chinese Information Processing Society of China.
- Cite (Informal):
- 半监督跨领域语义依存分析技术研究(Semi-supervised Domain Adaptation for Semantic Dependency Parsing) (Mao et al., CCL 2020)
- PDF:
- https://preview.aclanthology.org/ingest-acl-2023-videos/2020.ccl-1.73.pdf