CCL23-Eval 任务7系统报告:基于序列标注和指针生成网络的语法纠错方法(System Report for CCL23-Eval Task 7:A Syntactic Error Correction Approach Based on Sequence Labeling and Pointer Generation Networks)

Youren Yu (于右任), Yangsen Zhang (张仰森), Guanguang Chang (畅冠光), Beibei Gao (高贝贝), Yushan Jiang (姜雨杉), Tuo Xiao (肖拓)


Abstract
“针对当前大多数中文语法纠错模型存在错误边界识别不准确以及过度纠正的问题,我们提出了一种基于序列标注与指针生成网络的中文语法纠错模型。首先,在数据方面,我们使用了官方提供的lang8数据集和历年的CGED数据集,并对该数据集进行了繁体转简体、数据清洗等操作。其次,在模型方面,我们采用了ERNIE+Global Pointer的序列标注模型、基于ERNIE+CRF的序列标注模型、基于BART+指针生成网络的纠错模型以及基于CECToR的纠错模型。最后,在模型集成方面,我们使用了投票和基于ERNIE模型计算困惑度的方法,来生成最终预测结果。根据测试集的结果,我们的乃乏乍指标达到了48.68,位居第二名。”
Anthology ID:
2023.ccl-3.26
Volume:
Proceedings of the 22nd Chinese National Conference on Computational Linguistics (Volume 3: Evaluations)
Month:
August
Year:
2023
Address:
Harbin, China
Editors:
Maosong Sun, Bing Qin, Xipeng Qiu, Jing Jiang, Xianpei Han
Venue:
CCL
SIG:
Publisher:
Chinese Information Processing Society of China
Note:
Pages:
230–238
Language:
Chinese
URL:
https://aclanthology.org/2023.ccl-3.26
DOI:
Bibkey:
Cite (ACL):
Youren Yu, Yangsen Zhang, Guanguang Chang, Beibei Gao, Yushan Jiang, and Tuo Xiao. 2023. CCL23-Eval 任务7系统报告:基于序列标注和指针生成网络的语法纠错方法(System Report for CCL23-Eval Task 7:A Syntactic Error Correction Approach Based on Sequence Labeling and Pointer Generation Networks). In Proceedings of the 22nd Chinese National Conference on Computational Linguistics (Volume 3: Evaluations), pages 230–238, Harbin, China. Chinese Information Processing Society of China.
Cite (Informal):
CCL23-Eval 任务7系统报告:基于序列标注和指针生成网络的语法纠错方法(System Report for CCL23-Eval Task 7:A Syntactic Error Correction Approach Based on Sequence Labeling and Pointer Generation Networks) (Yu et al., CCL 2023)
Copy Citation:
PDF:
https://aclanthology.org/2023.ccl-3.26.pdf