인공신경망 특허 기계번역 성능에 관한 연구 - Patent Translate와 WIPO Translate 한영 번역 결과물의 누락과 통사 오류 분석을 중심으로 -

이지은,최효은 이화여자대학교 통역번역연구소 2022 T&I review Vol.12 No.2

Jieun Lee and Hyoeun Choi (2022). A study on the quality of patent neural machine translation: A comparison of omission and syntactic errors in the Korean-English translations by patent-specialized Patent Translate and WIPO Translate. This paper aims to evaluate the quality of patent translations produced by the two patent-specialized machine translation engines, EPO’s Patent Translate and WIPO’s WIPO Translate. For manual evaluation, four experienced patent translators or patent translation service managers evaluated the quality of 106 English sentences from the translations of 30 Korean patent abstracts by the two MT engines. In the automatic evaluation, Patent Translate slightly outperformed WIPO Translate, whereas in the manual evaluation WIPO Translate outperformed Patent Translate. According to the error annotations provided by the evaluators, WIPO Translate produced more omission errors than Patent Translate but handled the complex syntax of the source text better, while Patent Translate produced more syntactic errors than WIPO Translate. The results indicate that in automatic evaluation, MT outputs with fewer omissions were rated higher, while in manual evaluation, comprehensible and accurate syntactic structures appeared to determine the overall quality evaluation. (Ewha Womans University, Korea)

A Comparative Study of AI translation - Google Translate and Papago

Junsuk Park 한국중원언어학회 2022 언어학연구 Vol.- No.65

Since the mid 2010s, there have been a number of studies on the reliability of the translation capability of AI translation programs. Shin (2017) notes how machine translation systems have developed from a translation researcher’s perspective. Johnson et al. (2016) proposes a simple solution to use a single Neural Machine Translation (NMT) model to translate between multiple languages. Kim (2017) discusses how human spiritual aspect can be kept intact in the AI technology-dominated society. Kwak et al. (2020) conducted a study on AI translation focusing on proverb translation. Besides, Bahdanau et al. (2015) and Chung (2020) conducted research on machine translation. Following Park (2017, 2018)’s seminal research on the accuracy of the translation capability of Google Translate, this study investigates whether Google Translate and Papago, a translation program developed in Korea, have acquired native-like ability in translating sentences with semantically ambiguous Korean expressions such as –ya hayssta (had to vs was supposed to), -su issessta (could vs was able to), construction involving negation and sayngkakhata (think), mitta (believe), nukkita (feel) + embedded clause, and –cul alassta (knew or thought).

Machine Translation versus Human Translation: The Case of English-to-Chinese Translation of Relative Clauses

Zhuqing Wang,이승아 한국언어정보학회 2018 언어와 정보 Vol.22 No.1

Zhuqing Wang and Seung-Ah Lee. 2018. Machine Translation versus Human Translation: The Case of English-to-Chinese Translation of Relative Clauses. Language and Information 22.1, 175-201. This paper compares translations done by professional human translators (as manifested in the Babel English-Chinese Parallel Corpus) and machine translation sproduced by Google Translate (with English as the source language and Simplified Chinese as the target language). Capitalizing on the branching direction of these two languages, we investigated translation strategies of restrictive relative clauses by focusing on which-relatives. The specific objectives of the study were to (a) extract relevant data from a parallel corpus; (b) assess machine translation quality; and (c) highlight the similarities and dissimilarities between corpus outputs and machine translation outputs. Of the 147 test materials, 115 Google Translate outputs (78.2%) were rated as ‘successful’ or ‘acceptable’. No significant differences were found between the degree of success in machine translations and linguistic factors (e.g., the active or passive voice in relativeclauses). This finding confirms that linguistic knowledge is not required when using statistical machine translation (SMT) such as Google Translate. Another noteworthy finding was that Google Translate did not use the least frequently occurring translation strategy in the parallel corpus. This is not surprising given that SMT systems greatly rely on parallel corpora for training statistical translation models.

신경망 기계번역 내 젠더 문제 고찰 연구 – 네이버 파파고와 구글 번역을 중심으로 –

지윤주 한국통역번역학회 2021 통역과 번역 Vol.23 No.1

As a case study, this paper presents an investigation on gender bias in neural machine translation with Naver Papago and Google Translate. In spite of the remarkable progress in machine translation system, there are remaining challenging tasks, especially the gender bias in machine translation and processing itself. This kind of phenomenon threatens the fairness of a translation system and runs the risk of amplifying biases in machine translation. Under the circumstances, this study examines gender bias in neural machine translation system with Google Translate and Naver Papago. As for the gender bias evaluation in two machine translation, this paper makes English-Korean and Korean-English language sets using a lexicon set of adjectives. Also, this study purposes to compare Naver Papago and Google Translate. Upon the qualitative examination, the findings revealed that Naver Papago and Google Translate are significantly prone to gender-biased translation and biased results against women are much more frequent. Also, Naver Papagoseems to be more abusive against women. On the other hand, Google Translate seems to be more balanced than Naver Papago, using gender-neutral vocabulary and structure. Moreover, Naver papago and Google Translate exhibit a tendency toward male defaults typically associated with unbalanced gender distribution or stereotypes.

명대 백화단편소설의 번역과 전유

김소정(Kim, So-Jung) 대한중국학회 2018 중국학 Vol.62 No.-

본고는 명대 백화단편소설이 18세기에 최초로 영어로 번역된 이후 20세기 전반기까지 약 200년 동안 영어권 독자들에게 수용되어간 상황에 대해 고찰하였다. 명대 백화단편소설은 서양인 선교사의 선교활동과 유럽의 제국주의적 확장정책이 시작될 무렵에 번역되기 시작하여, 19세기와 20세기 동안 지속적으로 번역되어왔다. 본고에서는 번역본이 나온 시간적 순서에 따라 번역의 맥락과 특징 그리고 전유 양상 등을 중심으로 서술하였다. 18세기에 나온 최초의 번역은 불역본을 통한 간접번역이었으며, 19세기에 들어와서야 비로소 중국원전에서 직접 번역하기 시작했고 원문에 충실한 번역태도가 나타났다. 19세기까지 번역의 주체는 모두 서양인이었으며, 발표지면은 대부분 대중적인 신문잡지였고, 출판지역은 영국 런던, 오스트레일리아 세일, 일본 요코하마, 스코틀랜드 에딘버러, 중국의 홍콩, 광동, 상해 등 글로벌 국제도시에 두루 걸쳐있었다. 20세기에 들어서면서 중국인 번역가가 비로소 출현하였고 앤솔로지 형식의 단행본이 많이 출간되었으며 번역가의 번역의도에 따라 명대 백화 단편소설은 충실하게 또는 자유롭게 번역되었다. This article examines the situation in which vernacular short stories from Ming dynasty have been translated into english for about 200 years from the 18th century to the first half of the 20th century. Vernacular short stories from Ming dynasty began to be translated at the beginning time of missionary activities of Western missionaries and European imperialist expansionism. Since the first translation appeared in the 18th century, it has been continuously translated during the 19th and 20th centuries. As a result, Vernacular short stories from Ming dynasty was appropriated by the times that encountered unexpected Western readers. This article aims to describe the context of translation, translation strategy and appropriation aspects of the English translation version in chronological order. In the eighteenth century, the first translations were published through indirect translations based on French translations. Since the 19th century, a direct translation based on the Chinese original text appeared, the principle of faithful translation was emerged. In the nineteenth century, all translators were Westerners, and most of the translations were enclosed in popular newspaper or magazine, and the publishing areas spanned global cities such as London in England, Sail in Australia, Yokohama in Japan, Edinburgh in Scotland, Hong Kong, Guangdong and Shanghai in China. By the 19th century, Chinese Vernacular short stories has been widely spread to global cities. Compared to Western translators, Chinese translators did not appear until the 20th century. In the 20th century, the translators preferred direct translation rather than indirect translation, and many anthology-type books were published. According to the translation aim of the translator, domestication or foreignization strategy in translation was determined. Under the conditions of the target language culture, the extent of appropriation has often increased.

명대 백화단편소설의 번역과 전유 ― 18C-20C 전반기 영역본(英譯本)을 중심으로

김소정 대한중국학회 2018 중국학 Vol.62 No.-

This article examines the situation in which vernacular short stories from Ming dynasty have been translated into english for about 200 years from the 18th century to the first half of the 20th century. Vernacular short stories from Ming dynasty began to be translated at the beginning time of missionary activities of Western missionaries and European imperialist expansionism. Since the first translation appeared in the 18th century, it has been continuously translated during the 19th and 20th centuries. As a result, Vernacular short stories from Ming dynasty was appropriated by the times that encountered unexpected Western readers. This article aims to describe the context of translation, translation strategy and appropriation aspects of the English translation version in chronological order. In the eighteenth century, the first translations were published through indirect translations based on French translations. Since the 19th century, a direct translation based on the Chinese original text appeared, the principle of faithful translation was emerged. In the nineteenth century, all translators were Westerners, and most of the translations were enclosed in popular newspaper or magazine, and the publishing areas spanned global cities such as London in England, Sail in Australia, Yokohama in Japan, Edinburgh in Scotland, Hong Kong, Guangdong and Shanghai in China. By the 19th century, Chinese Vernacular short stories has been widely spread to global cities. Compared to Western translators, Chinese translators did not appear until the 20th century. In the 20th century, the translators preferred direct translation rather than indirect translation, and many anthology-type books were published. According to the translation aim of the translator, domestication or foreignization strategy in translation was determined. Under the conditions of the target language culture, the extent of appropriation has often increased. 본고는 명대 백화단편소설이 18세기에 최초로 영어로 번역된 이후 20세기 전반기까지 약 200년 동안 영어권 독자들에게 수용되어간 상황에 대해 고찰하였다. 명대 백화단편소설은 서양인 선교사의 선교활동과 유럽의 제국주의적 확장정책이 시작될 무렵에 번역되기 시작하여, 19세기와 20세기 동안 지속적으로 번역되어왔다. 본고에서는 번역본이 나온 시간적 순서에 따라 번역의 맥락과 특징 그리고 전유 양상 등을 중심으로 서술하였다. 18세기에 나온 최초의 번역은 불역본을 통한 간접번역이었으며, 19세기에 들어와서야 비로소 중국원전에서 직접 번역하기 시작했고 원문에 충실한 번역태도가 나타났다. 19세기까지 번역의 주체는 모두 서양인이었으며, 발표지면은 대부분 대중적인 신문잡지였고, 출판지역은 영국 런던, 오스트레일리아 세일, 일본 요코하마, 스코틀랜드 에딘버러, 중국의 홍콩, 광동, 상해 등 글로벌 국제도시에 두루 걸쳐있었다. 20세기에 들어서면서 중국인 번역가가 비로소 출현하였고 앤솔로지 형식의 단행본이 많이 출간되었으며 번역가의 번역의도에 따라 명대 백화단편소설은 충실하게 또는 자유롭게 번역되었다.

An Analysis of Korean-English Translation Errors in Google Translate

이정화,차경환 언어과학회 2019 언어과학연구 Vol.0 No.89

The purpose of this study is to identify types of errors that occur when using Google Translate and to suggest more effective ways to benefit from machine translation translators. The topics in the 30 news articles dealt with health and diet. The researchers directed the error analysis by comparing the English translation of the original Korean text with the eleven categories and their sub-categories. Among those translated by the Google Translate Program, according to our research, the number of sentences that contained errors was 365 (67.2%) and that of correct sentences was 178 (32.8%). Sentence structure errors, the highest of the eleven categories, recorded 172 (21.31%) cases, followed by nouns with 143 (17.72%), and verbs with 119 (14.75%). The findings emphasize the need to learn how to use machine translators effectively. Finally, Google Translate users need to avoid using complex sentences and vague words when translating Korean text into English.

원천텍스트 난이도와 기계번역 품질에 대한 고찰 - 구글 한영 법령번역 사례를 중심으로

이지은,최효은 한국언어연구학회 2023 언어학연구 Vol.28 No.1

This paper attempts to investigate the relationship between the source text difficulty and the quality of machine translation output. We examined the results of human evaluation of the Korean-English statutory translation by Google Translate and the source text difficulty measured in terms of lexical and terminological difficulty and syntactic complexity. The translations of two different statutes revealed varying correlation patterns examined in terms of fluency, accuracy, terminology and others. The results demonstrate that the two source texts’ syntactic complexity had a negative correlation with the machine translation quality evaluation. In other words, Google Translate was found to be weak at translating long and syntactically complex sentences. On the other hand, lexical and terminological difficulty did not have any correlation with the results of human evaluation despite the fact that statutes contain a lot of difficult words. Overall, the source text difficulty had a strong negative correlation with the human evaluation results, which was generally supported by the evaluators’ comments on fluency and accuracy errors.

An Analysis of Korean-English Translation Errors in Google Translate

Jeong-Hwa Lee(이정화),Kyung-Whan Cha(차경환) 언어과학회 2019 언어과학연구 Vol.0 No.89

The purpose of this study is to identify types of errors that occur when using Google Translate and to suggest more effective ways to benefit from machine translation translators. The topics in the 30 news articles dealt with health and diet. The researchers directed the error analysis by comparing the English translation of the original Korean text with the eleven categories and their sub-categories. Among those translated by the Google Translate Program, according to our research, the number of sentences that contained errors was 365 (67.2%) and that of correct sentences was 178 (32.8%). Sentence structure errors, the highest of the eleven categories, recorded 172 (21.31%) cases, followed by nouns with 143 (17.72%), and verbs with 119 (14.75%). The findings emphasize the need to learn how to use machine translators effectively. Finally, Google Translate users need to avoid using complex sentences and vague words when translating Korean text into English.

법률 특화 번역엔진 성능 평가 - 한영 계약서 번역을 중심으로 -

이준호 이화여자대학교 통역번역연구소 2022 T&I review Vol.12 No.1

This study aims to examine the possibility to enhance the performance of a machine translation engine in Korean-to-English legal translation. With the rise of neural machine translation, translation scholars quickly responded to investigate the output and limitations of machine translation. However, there has been little research to date on enhancing the quality of raw machine translation itself. As the legal field generates vast demand for machine translation post-editing, any improvement in machine translation engine performance can be of significant contribution. Hence, this paper will introduce an experimental machine translation engine trained with a bilingual legal corpus and verify whether it is possible to create a domain-specific engine that can outperform a freely available general purpose translation service. To that end, the output from the experimental engine and Google Translate will be compared and analyzed in terms of accuracy, fluency, usability, and essential changes. Last but not least, this paper will examine the potential contribution that translators can make and areas of collaboration between translation scholars and translation engine developers. (Hankuk University of Foreign Studies, Korea)