shibing624 commited on
Commit
036411f
1 Parent(s): ca9ef2f

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +134 -0
README.md CHANGED
@@ -1,3 +1,137 @@
1
  ---
2
  license: apache-2.0
 
 
 
 
 
 
 
3
  ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
  ---
2
  license: apache-2.0
3
+ language:
4
+ - zh
5
+ tags:
6
+ - text-correction
7
+ pretty_name: CSC
8
+ task_categories:
9
+ - text-generation
10
  ---
11
+
12
+ # Dataset Card for CSC
13
+
14
+ - **Repository:** https://github.com/shibing624/pycorrector
15
+
16
+ ## Dataset Description
17
+
18
+ Chinese Spelling Correction (CSC) is a task to detect and correct misspelled characters in Chinese texts.
19
+
20
+ CSC is challenging since many Chinese characters are visually or phonologically similar but with quite different semantic meanings.
21
+
22
+ SIGHAN+Wang271K中文纠错数据集(27万条),是通过原始SIGHAN13、14、15年数据集和Wang271K数据集格式转化后得到,json格式,带错误字符位置信息,SIGHAN为test.json。
23
+
24
+
25
+ ### Original Dataset Summary
26
+
27
+ test.json 为 **SIGHAN数据集**, 包括SIGHAN13 14 15,来自 [官方csc.html](http://nlp.ee.ncu.edu.tw/resource/csc.html) ,大小:339K。
28
+ train.json 和 dev.json 为 **Wang271K数据集**,包括 Wang271K ,来自 [Automatic-Corpus-Generation dimmywang提供](https://github.com/wdimmy/Automatic-Corpus-Generation/blob/master/corpus/train.sgml) ,大小: 93M 。
29
+
30
+ ### Supported Tasks and Leaderboards
31
+
32
+ The Alpaca dataset designed for instruction training pretrained language models.
33
+
34
+ ### Languages
35
+
36
+ The data in CSC are in Chinese.
37
+
38
+ ## Dataset Structure
39
+
40
+ ### Data Instances
41
+
42
+ An example of "train" looks as follows:
43
+
44
+ ```json
45
+ {
46
+ "id": "B2-4029-3",
47
+ "original_text": "晚间会听到嗓音,白天的时候大家都不会太在意,但是在睡觉的时候这嗓音成为大家的恶梦。",
48
+ "wrong_ids": [
49
+ 5,
50
+ 31
51
+ ],
52
+ "correct_text": "晚间会听到噪音,白天的时候大家都不会太在意,但是在睡觉的时候这噪音成为大家的恶梦。"
53
+ }
54
+ ```
55
+
56
+
57
+ ### Data Fields
58
+
59
+ 字段解释:
60
+ - id:唯一标识符,无意义
61
+ - original_text: 原始错误文本
62
+ - wrong_ids: 错误字的位置,从0开始
63
+ - correct_text: 纠正后的文本
64
+
65
+
66
+ ### Data Splits
67
+
68
+ | | train | dev | test |
69
+ |---------------|------:|--:|--:|
70
+ | CSC | 200K | 71K | 4K |
71
+
72
+ ## Dataset Creation
73
+
74
+ ### Curation Rationale
75
+
76
+ [More Information Needed]
77
+
78
+ ### Source Data
79
+
80
+ #### Initial Data Collection and Normalization
81
+
82
+ [More Information Needed]
83
+
84
+ #### Who are the source language producers?
85
+
86
+ [More Information Needed]
87
+
88
+ ### Annotations
89
+
90
+ #### Annotation process
91
+
92
+ [More Information Needed]
93
+
94
+ #### Who are the annotators?
95
+
96
+ [More Information Needed]
97
+
98
+ ### Personal and Sensitive Information
99
+
100
+ [More Information Needed]
101
+
102
+ ## Considerations for Using the Data
103
+
104
+ ### Social Impact of Dataset
105
+
106
+
107
+ ### Discussion of Biases
108
+
109
+ [More Information Needed]
110
+
111
+ ### Other Known Limitations
112
+
113
+ ## Additional Information
114
+
115
+ ### Dataset Curators
116
+
117
+ [More Information Needed]
118
+
119
+ ### Licensing Information
120
+
121
+ The dataset is available under the Apache 2.0.
122
+
123
+
124
+ ### Citation Information
125
+
126
+ ```latex
127
+ @misc{Xu_Pycorrector_Text_error,
128
+ title={Pycorrector: Text error correction tool},
129
+ author={Xu Ming},
130
+ year={2021},
131
+ howpublished={\url{https://github.com/shibing624/pycorrector}},
132
+ }
133
+ ```
134
+
135
+ ### Contributions
136
+
137
+ [shibing624](https://github.com/shibing624) 整理并上传