초록 열기/닫기 버튼
This study is an attempt to make productive recommendations for the improvement of translation quality in automated translation. With the advent of automated translation in 1949, the field of machine translation has made remarkable progress. Despite its advancement, however, translation quality has failed to meet the expectations of its users. In this sense, it is timely and appropriate to seek ways to improve the translation quality of automated translation services as the user base for machine translation has been rapidly expanding. Against this backdrop, as a contribution towards finding ways to improve translation quality in machine translation, this study explores the correlation between translation quality and the units of translations in automated translation by using “Google Translate”, one of the most popular statistic-based machine translation tools. A set of Korean product information texts, which are composed of a total of 23 smartphone models compiled from the Samsung Galaxy website, are compared and analyzed with their parallel texts translated by Google Translate into English and French. To clearly demonstrate the correlation between units of translation and translation quality in the original and translated texts are used as analytical criteria, which are divided into three categories: functional units, semantic units, and dialectical units, which were first proposed by Vinay & Darbelnet (1958). And based on the results of the analysis, this study attempts to make constructive recommendations for translation quality improvement in statistic-based machine translation models.
