본문 바로가기
  • Home

Error Correction Using Korean-Chinese Machine Translation and Post-Editing Productivity

  • Journal of Chinese Language and Literature
  • 2022, (90), pp.197-217
  • DOI : 10.15792/clsyn..90.202208.197
  • Publisher : Chinese Literary Society Of Yeong Nam
  • Research Area : Humanities > Chinese Language and Literature
  • Received : July 10, 2022
  • Accepted : August 13, 2022
  • Published : August 30, 2022

Yeonok Hong 1

1세종대학교

Accredited

ABSTRACT

Efforts to use machine translation for foreign language learning are currently conducted by researchers and educators who study English. On the other hand, attempts to use machine translation have been conducted in the Chinese language education, but discussions on specific teaching methods are still in the rudimentary stage. Therefore, this study attempts to propose a teaching method that can use machine translation effectively in the field of Chinese teaching using an error correction teaching method with machine translation. The types of errors in Chinese translation that frequently appear in Korean learners, are considered, and the results are compared and analyzed by extracting their translation from neural machine translation. In addition, to modify the results output from machine translation and to use them as high-quality translation as a collaboration with advanced science technology and humans, the study attempts to post-edit each other's results, and to analyze the Chinese learning effect using machine translation. As a result of the analysis, humans and machines are able to improve the results of translation and accumulate accurate learning data in the process of correcting each other's errors. The accumulated results will increase the ability of humans to speak a fluent language, and machine translation will consist of accurate function data essential in deep learning systems and will be used as re-learning data to increase translation accuracy.

Citation status

* References for papers published after 2023 are currently being built.

This paper was written with support from the National Research Foundation of Korea.