Thai Deaf Corpus

Citation Author(s):
Supachan
Traitruengsakul
Department of Computer Engineering, Faculty of Engineering, Chulalongkorn University
Ekapol
Chuangsuwanich
Department of Computer Engineering, Faculty of Engineering, Chulalongkorn University
Submitted by:
Supachan Traitr...
Last updated:
Fri, 07/26/2024 - 12:06
DOI:
10.21227/53w2-1k42
Data Format:
TXT
License:
Creative Commons Attribution
226 Views
Categories:
Other
Keywords:
Deaf Writing, Grammatical Error Correction, artificial intelligence; machine learning; natural language processing;
0
0 ratings - Please login to submit your rating.
ACCESS DATASET CITE SHARE/EMBED

Abstract 

The Thai Deaf Corpus (TDC) is constructed from a writing activity where deaf students randomly select picture words using the image picker wheel, then write sentences corresponding to these words on the writing sheet. The sentences are transcribed and corrected manually to create the TDC.

  • It contains 22,719 sentences written by deaf students, with their corresponding corrections separated by "|||".
  • For example, sentence x may have one or more possible corrections such as sentence y1, sentence y2, and so on. We may get sentence pairs: sentence x ||| sentence y1, sentence x ||| sentence y2, ...
Instructions: 

In TDC, each line contains (1) an original sentence written by deaf students, (2) "|||" separation, and (3) its correction by native speakers.

Funding Agency: 
Foundation for the Deaf under the Royal Patronage of Her Majesty the Queen

Dataset Files

LOGIN TO ACCESS DATASET FILES

 

QUESTIONS?