基于ChineseBert的中文拼写纠错方法
崔凡, 强继朋, 朱毅, 李云
Chinese spelling correction method based on ChineseBert
Fan Cui, Jipeng Qiang, Yi Zhu, Yun Li
表1
实验中使用的数据集统计
Table 1
Dataset statistics used in experiments
Train Set
#Sent
Avg.Length
#Errors
(wang)
271329
44.4
271329
SIGHAN2013
700
49.2
350
SIGHAN2014
3435
49.7
3432
SIGHAN2015
2339
30.0
2339
Test Set
#Sent
Avg.Length
#Errors
SIGHAN2013
1000
74.1
996
SIGHAN2014
1062
50.1
529
SIGHAN2015
1100
30.5
550