Abstract
“本文介绍了参赛系统在第三届中文空间语义理解评测(SpaCE2023)采用的技术路线:面向空间语义异常识别任务提出了抽取方法,并结合生成器进一步完成了空间语义角色标注任务,空间场景异同判断任务则使用了大语言模型生成。本文进一步探索了大语言模型在评测数据集上的应用,发现指令设计是未来工作的重点和难点。参赛系统的代码和模型见https://github.com/ShacklesLay/Space2023。”- Anthology ID:
- 2023.ccl-3.13
- Volume:
- Proceedings of the 22nd Chinese National Conference on Computational Linguistics (Volume 3: Evaluations)
- Month:
- August
- Year:
- 2023
- Address:
- Harbin, China
- Venue:
- CCL
- SIG:
- Publisher:
- Chinese Information Processing Society of China
- Note:
- Pages:
- 139–149
- Language:
- Chinese
- URL:
- https://aclanthology.org/2023.ccl-3.13
- DOI:
- Cite (ACL):
- ChenKun Tan, XianNian Hu, and XinPeng Qiu. 2023. CCL23-Eval任务4系统报告:基于深度学习的空间语义理解(System Report for CCL23-Eval Task4:Spatial Semantic Understanding Based on Deep Learning.). In Proceedings of the 22nd Chinese National Conference on Computational Linguistics (Volume 3: Evaluations), pages 139–149, Harbin, China. Chinese Information Processing Society of China.
- Cite (Informal):
- CCL23-Eval任务4系统报告:基于深度学习的空间语义理解(System Report for CCL23-Eval Task4:Spatial Semantic Understanding Based on Deep Learning.) (Tan et al., CCL 2023)
- PDF:
- https://preview.aclanthology.org/remove-xml-comments/2023.ccl-3.13.pdf