• 비교문화연구
• /
• 제19권
• /
• pp.251-270
• /
• 2010

The subject of the paper is to discern the validity of the translation theory put forward by the ESIT(Ecole Sup?rieur d'Interpr?tes et de Tranducteurs, Universit? Paris III) and how it differentiates from the other translation theories. First, the paper will analyze the theoretical aspects put forward by examining the equivalence that may be discerned between the french and korean translation in relation to the original english text that is being translated. Employing the equivalence in translation may shed new insights into the unterminable discussions we witness today between the literal translation and the free translation. Contrary to the formal equivalence the dynamic equivalence by Nida suggests that the messages retain the same meanings whether it be the original or a translated text to the/for the reader. In short, the object of the dynamic equivalence is to identify the closest equivalence to the suggested source language. The concept of correspondence and equivalence defined by theoriticians of translation falls to the domain of dynamic equivalence suggested by Nida. In translation theory the domain of usage of language and the that of discourse is denoted separately. by usage one denotes the translation through symbols that make up language itself. In contrast to this, the discourse is suggestive of defining the newly created expressions which may be denoted as being a creative equivalence which embodies the original message for the singular situation at hand. The translator will however find oneself incorporating the two opposing theories in translating. Translation falls under the criteria of text and not of language, thus one cannot regulate or foresee any special circumstances that may arise in translation of discourse, the translation to reflect this condition should always be delimited. All other translation should be subject to translation by equivalence. The interpretation theory in translation (of ESIT) in effect is relative to both the empirical and philosophical approach and is suggestive of new perspective in translation. In conclusion, the above suggested translation theory is different from the skopos theory and the polysystem theory in that it only takes in to account the elements that are in close relation to the original text, and also that it was developed for educational purposes opening new perspectives in the domain of translation theories.

• 한국멀티미디어학회논문지
• /
• 제25권10호
• /
• pp.1535-1545
• /
• 2022

Sign language has completely different meaning depending on the direction of the hand or the change of facial expression even with the same gesture. In this respect, it is crucial to capture the spatial-temporal structure information of each movement. However, sign language translation studies based on Sign2Gloss2Text only convey comprehensive spatial-temporal information about the entire sign language movement. Consequently, detailed information (facial expression, gestures, and etc.) of each movement that is important for sign language translation is not emphasized. Accordingly, in this paper, we propose Spatial-temporal Keypoints Centered Sign2Gloss2Text Translation, named STKC-Sign2 Gloss2Text, to supplement the sequential and semantic information of keypoints which are the core of recognizing and translating sign language. STKC-Sign2Gloss2Text consists of two steps, Spatial Keypoints Embedding, which extracts 121 major keypoints from each image, and Temporal Keypoints Embedding, which emphasizes sequential information using Bi-GRU for extracted keypoints of sign language. The proposed model outperformed all Bilingual Evaluation Understudy(BLEU) scores in Development(DEV) and Testing(TEST) than Sign2Gloss2Text as the baseline, and in particular, it proved the effectiveness of the proposed methodology by achieving 23.19, an improvement of 1.87 based on TEST BLEU-4.

• 영어영문학
• /
• 제57권2호
• /
• pp.217-235
• /
• 2011

On recognizing the significance of Walter Benjamin's "The Task of a Translator" in recent discourses of postcolonial cultural translation, this essay examines the creative postcolonialist appropriations of Benjamin's theory of translation and their political implications. In an effort to dismantle the imperialist political hierarchy between the West and the non-West, modernity and its "primitive" others, which has been the operative premise of the traditional translation studies and anthropology, newly emergent discourses of cultural translation actively adopts Benjamin's notion of translation that does not prioritize the original text's claim on authenticity. Benjamin theorizes each text-translation as well as the original-as an incomplete representation of the pure language. Eschewing formalistic views propounded by deconstructionist critics like Paul de Man, who tend to regard Benjamin's notion of the untranslatable purely in terms of the failure inherent in the language system per se, such postcolonialist critics as Tejaswini Niranjana, Rey Chow, and Homi Bhabha, each in his/her unique way, recuperate the significatory potential of historicity embedded in Benjamin's text. Their further appropriation of the concept of the "untranslatable" depends on a radically political turn that, instead of focusing on the failure of translation, salvages historical as well as cultural potentiality that lies between disparate cultural entities, signifying differences, or disjunctures, that do not easily render themselves to existing systems of representation. It may therefore be concluded that postcolonial discourses on cultural translation of Niranhana, Chow, and Bhabha, inspired by Benjamin, each translate the latter's theory into highly politicized understandings of translation, and this leads to an extensive rethinking of the act of translation itself to include all forms of cultural exchange and communicative activities between cultures. The disjunctures between these discourses and Benjamin's text, in that sense, enable them to form a sort of theoretical constellation, which aspires to an impossible yet necessary utopian ideal of critical thinking.

• 한국정보처리학회:학술대회논문집
• /
• 한국정보처리학회 2009년도 추계학술발표대회
• /
• pp.239-240
• /
• 2009

We live in a flat world in which globalization fosters communication, travel, and trade among more than 150 countries and thousands of languages. To surmount the barriers among these languages, translation is required; Speech-to-Speech translation will automate the process. Thanks to recent advances in Automatic Speech Recognition (ASR), Machine Translation (MT), and Text-to-Speech (TTS), one can now utilize a system to translate a speech of source language to a speech of target language and vice versa in affordable manner. The three phase process establishes that the source speech be transcribed into a (set of) text of the source language (ASR) before the source text is translated into the target text (MT). Finally, the target speech is synthesized from the target text (TTS).

• 스마트미디어저널
• /
• 제12권1호
• /
• pp.32-46
• /
• 2023

Natural Language Processing (NLP) has grown tremendously in recent years. Typically, bilingual, and multilingual translation models have been deployed widely in machine translation and gained vast attention from the research community. On the contrary, few studies have focused on translating between spoken and sign languages, especially non-English languages. Prior works on Sign Language Translation (SLT) have shown that a mid-level sign gloss representation enhances translation performance. Therefore, this study presents a new large-scale Korean sign language dataset, the Museum-Commentary Korean Sign Gloss (MCKSG) dataset, including 3828 pairs of Korean sentences and their corresponding sign glosses used in Museum-Commentary contexts. In addition, we propose a translation framework based on self-supervised learning, where the pretext task is a text-to-text from a Korean sentence to its back-translation versions, then the pre-trained network will be fine-tuned on the MCKSG dataset. Using self-supervised learning help to overcome the drawback of a shortage of sign language data. Through experimental results, our proposed model outperforms a baseline BERT model by 6.22%.

• 한국정보처리학회논문지
• /
• 제5권3호
• /
• pp.703-714
• /
• 1998

기존의 점자 부호 체계에서 한글문장을 점자문장으로 번역하는 작업은 1 : 1의 대응관계가 있으므로 아무런 문제를 야기시키지 않는다. 그러나 점자문장을 한글문장으로 번역하는 작업은 점자부호와 한글부호간의 1 : N의 대응관계로 인하여 피할 수 없는 오류를 파생시킨다. 분석 결과, 점자 부호 체계에서는 한글 초성과 숫자와의 충돌, 한글 종성과 문장부호와의 충돌, 한글 소괄호 열기와 소괄호 닫기 부호를 동일한 부호로 처리하는 것, 영문 열기 및 닫기 부호와 한글 종성과의 충돌 등이 번역 오류의 중요한 원인임을 알 수 있다. 본 논문에서는 이러한 충돌을 해소하기 위한 역점역 규칙을 제정하여, 일차적으로 역점역시에 발생하는 애매성을 앞, 뒤 문맥에 따라 해결할 수 있는 경우에는 원래의 점자 부호를 사용하도록 하고, 만일 이 방법으로 해결되지 않을 경우에는 역점역시에 충돌을 근본적으로 방지하는 확장 점자 부호를 사용하도록 하였다. 여러 가지 파일에 대한 실험 결과, 기존의 점자 부호체계에서의 오류발생율이 평균 25%이었음에 비하여 확장된 점자 부호체계에서는 번역에 따른 오류가 발견되지 않았다.

• ETRI Journal
• /
• 제45권1호
• /
• pp.18-27
• /
• 2023

We present an English-Korean speech translation corpus, named EnKoST-C. End-to-end model training for speech translation tasks often suffers from a lack of parallel data, such as speech data in the source language and equivalent text data in the target language. Most available public speech translation corpora were developed for European languages, and there is currently no public corpus for English-Korean end-to-end speech translation. Thus, we created an EnKoST-C centered on TED Talks. In this process, we enhance the sentence alignment approach using the subtitle time information and bilingual sentence embedding information. As a result, we built a 559-h English-Korean speech translation corpus. The proposed sentence alignment approach showed excellent performance of 0.96 f-measure score. We also show the baseline performance of an English-Korean speech translation model trained with EnKoST-C. The EnKoST-C is freely available on a Korean government open data hub site.

• 비교문화연구
• /
• 제36권
• /
• pp.241-270
• /
• 2014

The aim of this article is to review the traditional methodologies of teaching translation that concentrate on text-typologies and, as an alternative, to propose an eclectic multi-componential approach that involves a set of interdisciplinary skills with a view to improving the trainee translators' competences and skills. To this end, three approaches, namely a minimalist approach, a pre-transferring adjustment approach and a revision vs. editing approach are proposed to shift the focus of attention from teacher-centred approaches towards student-centred approaches. It has been shown that translator training programmes need to focus on improving the trainee translators' competences and skills, such as training them how to produce and select among the different versions they produce by themselves with justified confidence as quickly as they can (minimalist approach), adjust the original text semantically, syntactically and/or textually in a way that the source text supplely accommodates itself in the linguistic system of the target language (pre-transferring adjustment), and revise and edit others' translations. As the validity of the approach proposed relies partially on instructors' competences and skills in teaching translation, universities, particularly in the Arab world, need to invest in recruiting expert practitioners instead of depending mainly on bilingual teachers to teach translation.

• 한국언어정보학회:학술대회논문집
• /
• 한국언어정보학회 2002년도 Language, Information, and Computation Proceedings of The 16th Pacific Asia Conference
• /
• pp.147-156
• /
• 2002

Aligned parallel corpora have proved very useful in many natural language processing tasks, including statistical machine translation and word sense disambiguation. In this paper, we describe an alignment technique for extracting transfer mapping from the parallel corpus. During building our system and data collection, we observe that there are three types of translation approaches can be used. We especially focuses on Traditional Chinese and Simplified Chinese text lexical translation and a method for extracting transfer mappings for machine translation.

• 한국문헌정보학회지
• /
• 제43권1호
• /
• pp.313-332
• /
• 2009

교차언어 문서 범주화(CLTC)는 다른 언어로 된 학습집단을 이용하여 문헌을 자동 분류할 수 있다. 이 연구는 KTSET으로부터 CLTC에 적합한 실험문헌집단을 추출하고, 기계 번역기를 이용하여 가능한 여러 CLTC 방법의 분류 성능을 비교하였다. 분류기는 SVM 분류기를 이용하였다. 실험 결과, CLTC 중에 다국어 학습방법이 가장 좋은 분류 성능을 보였으며, 학습집단 번역방법, 검증집단 번역방법 순으로 분류 성능이 낮아졌다. 하지만 학습집단 번역방법이 기계번역 측면에서 효율적이며, 일반적인 환경에 쉽게 적용할 수 있고, 비교적 분류 성능이 좋아 CLTC 방법 중에서 가장 높은 이용 가능성을 보였다. 한편 CLTC에서 기계번역을 이용하였을 때 번역과정에서 발생하는 자질축소나 주제적 특성이 없는 자질로의 번역으로 인해 성능 저하를 가져왔다.