High-quality Translation Research Articles

The translator, an endangered species?

The first public machine translation experiment took place in 1954. Led by researchers from IBM and Georgetown University in Washington, DC, it was destined to make possible high-quality automatic translation from Russian into English in a few years. Since this first attempt, claims that machines could soon be replacing translators became usual. In 2018, Microsoft announced that their Chinese to English news translations were of comparable quality to human translation. The paradox however, is that although translation systems are accessible to most people, the number of people working in the translation industry is higher than ever before – an approximate 600,000 people in the world. In this context, do professionals really have a reason to worry?

机器翻译辅助的中蒙、维汉语音翻译数据集子集

At present, there are few public datasets for speech translation, especially those between Chinese and other low-resource languages. The development of end-to-end speech translation is limited by resources. In light of the research idea of international speech translation datasets, in this paper, we used the public speech recognition datasets (AISHELL and THUYG-20) to convert them into speech translation datasets through machine translation. After data processing, they were reviewed and verified by experts, so as to obtain high-quality speech translation datasets. The dataset includes Chinese-Mongolian speech translation dataset and Uygur-Chinese speech translation dataset, and the audio sampling rate is 16 kHz. The Chinese-Mongolian speech translation subset contains 1,919 items with a size of 238 MB. The Uygur-Chinese speech translation subset contains 3,692 samples with a size of 652 MB. This dataset can be used for the research on end-to-end speech translation, and provide data support for exploring the speech translation between Chinese and minority languages. As the dataset has been reviewed and verified by experts, it can also be combined with speech recognition dataset to study machine translation.

English Live Caption in a Javanese Shadow Puppet Theatre: Systemic Functional Linguistics for Appraising the Beauty of Sagopi

In the field of simultaneous translation, live captioning has not been sufficiently touched by research. There have been shifts in translating characterization due to this being done in real-time and on the spot. It is critical that target readers or viewers get the same impression of a character's personality as portrayed in the original work. Consequently, translations of literary plays must be thoroughly checked for fidelity. The objective of the study is to undertake an in-depth analysis of the live captioning provided in "The Life Struggle of Sagopi" puppet theater. It compares the character portrayal of the protagonist, Sagopi, in the original play and the translation. Employing a content analysis technique, all of the utterances describing Sagopi's beauty were collected. As a method for analyzing the data, the theory of Appraisal was utilized. When the described Appraisal elements are identical to those in the original, the characterization would be considered equivalent. The analysis reveals that the live captions do not fully represent the portrayal of Sagopi. While the description of Sagopi's beauty in the shadow puppet performance is detailed, the English live caption does not reflect this. In the target language, deletions and shifts of appraisals indicate the changes. The lack of culturally relevant terminologies and the time lag between the puppeteer's oral utterances and the translator's typing speed during live captioning are the main causes of the problem. Furthermore, live captioning is time and space-constrained. Almost always, it is a condensed version of the oral source text. This demonstrates how difficult it is to achieve the same level of appraisal as the original. This is not to suggest that viewers have no right to a high-quality translation. Despite its shortcomings, the written text in the form of a live caption is regarded as the finest way for foreigners to enjoy a puppet show.

Обучение китайских студентов-русистов навыкам перевода: учёт принципов теории экотранслятологии (на примере рассказа А.П. Чехова «Зиночка»)

Introduction. The most important area of research in the methodology of teaching Russian as a foreign language is teaching translation to foreign students. The quality of translation depends not only on the degree of proficiency in Russian, but also on how well students can use their native language. This is also observed among Chinese students who have difficulty translating Russian-language texts into Chinese. Hypothesis: the method of teaching translation using the theory of “eco-translatology” developed by Professor Hu Genshen of Tsinghua University (China) is relevant. The purpose of the study is to prove the effectiveness of the application of the Hu Genshen's theory in the formation and development of literary translation skills in Chinese students. Materials and methods: the theoretical analysis of scientific and methodological literature, content analysis, comparative analysis, description and analysis of research results. A group of 20 Chinese third-year students of the "Philology" training course took part in the experiment. They performed tasks on the written translation of A.P. Chekhov's story “Zinochka”. Results: variants of translations of several fragments of this work are considered. The advantages and disadvantages of students' translations are noted; the significance of the main approach of the theory of eco-translatology – “adaptation and selection” – for teaching translation is evaluated. Conclusions. To obtain an adequate translation, it is necessary to deepen the work on the original text and the theory of translation. The teacher needs to develop students' ability to sense the emotional mood of the characters through specially organized work on this aspect. It is important for students to acquire the skills of taking into account the context, situational and stylistic features, to learn the structural differences of languages and strive to bring their knowledge of Chinese to a higher level, to develop vocabulary as a prerequisite for high-quality translation of texts of different speech styles.

A Study of Pre-editing Methods at the Lexical Level in the Process of Machine Translation

As the demand for translation increases, the role of machine translation in improving the efficiency of translation is increasingly prominent, but due to its inherent limitations, the translation quality of machine translation is not guaranteed. Pre-editing is one of the ways to enhance the quality of machine translated texts. With Google online translation as a tool, the effect of pre-editing on the quality of machine translation is studied by adopting the Bilingual Evaluation Understudy (BLEU) approach. The pre-editing methods of additions, omissions, replacing and terminology preprocessing based on the lexical level are proposed. The results show these methods can play a role in improving the quality of English-to-Chinese machine translations. In most cases, pre-editing at the lexical level is sufficient to generate high-quality machine translation output, but the effectiveness of pre-editing can be further improved when perspective-shifting is involved or when the target language involves contextually consistent verb or person usage.

A Model for Gujarati News Search Engine by Link Builder and Web Crawler Algorithms

The web serves a paradoxical role in language learning. Providing a fertile ground for neologisms and redefinitions, the web incubates a growing translation gap between one’s native tongue and his or her target foreign language. However, the web can also be used to bridge languages even as they change, because it contains a vast and ever-expanding set of human-translated web pages. Search engines that mine high-quality human translations can unlock the web’s pedagogical potential, and form the basis for next-generation language learning tools. So, it has become essential to develop crawlers in such a way that it can help in local language optimization with topic-search. Based on study of many research papers motivated to do research in Gujarati Search Engine. Moreover, proposed a new model for it.

A Model for Gujarati News Search Engine by Link Builder and Web Crawler Algorithms

The web serves a paradoxical role in language learning. Providing a fertile ground for neologisms and redefinitions, the web incubates a growing translation gap between one’s native tongue and his or her target foreign language. However, the web can also be used to bridge languages even as they change, because it contains a vast and ever-expanding set of human-translated web pages. Search engines that mine high-quality human translations can unlock the web’s pedagogical potential, and form the basis for next-generation language learning tools. So, it has become essential to develop crawlers in such a way that it can help in local language optimization with topic-search. Based on study of many research papers motivated to do research in Gujarati Search Engine. Moreover, proposed a new model for it.

Automatic Extraction of English-Chinese Translation Templates Based on Deep Learning

Translation templates are an important cause of knowledge in machine translation (MT) systems. Their quality and scale directly influence the performance of MT systems. How to obtain high-quality and efficient translation templates from corpora has become a hot topic in recent study. In this paper, a tree to String alignment template (TAT) based on syntactic structure is proposed. This template describes the alignment between the source language syntax tree and the target language string. The syntactic structure, a large number of construction tags, and variables are introduced into the template, which enables the syntactic model to deal with discontinuous phrases and has the ability of generalization. Templates can be used in syntactic statistics, case-based, and rule-based MT systems according to different decoders. ATTEBSC algorithm is a basic method to learn translation templates by comparing sentence pairs. It demands that sentence pairs be constructed in a precise comparison structure ahead of time, but there are no strict guidelines on how to do it. In this paper, we propose a method to calculate the specific comparison scheme using the longest common subsequence (LCS) and use the normalized LCS distance to screen sentences with high similarity and then use the ATTEBSC algorithm to automatically remove the template. Experiments show that this method is easy and effective, and many expensive templates can be learned.

513 Towards Obtaining One Billion Recordings Per Cubic Millimeter for the Validation of Focused Ultrasound Transducers: How Can Robust Systems Help With Translational Activities?

OBJECTIVES/GOALS: The first aim was to construct a controlled and high resolution FUS water tank characterization system with 1 micron step-sizes. The second aim was to create two unique standardized protocols for mapping the generated acoustic field from FUS transducers; protocol one maps the full 3D field while protocol two rapidly detects changes to the original plot. METHODS/STUDY POPULATION: To accomplish aim one, the focused ultrasound mapping platform was constructed with a water conditioning unit for water degassing and temperature control, a three-axis stage with 1 micron step-size capabilities, and a data plotting software. To measure the outcomes of aim one, the water temperature was monitored, and axis step sizes were measured through ten independent axis translation recordings. To accomplish aim two, FUS acquisitions were executed at different resolutions. For FUS localization at the cellular level, a 1-5 micron step size is required. Once the initial scan was performed, duplicate scans were executed to detect inherent perturbations or errors in the system. Once calculated, the best methods of detecting true changes to FUS signals are proposed. RESULTS/ANTICIPATED RESULTS: The FUS characterization system maintained water temperature and performed 1 micron step-sizes. While pre-existing platforms have demonstrated a resolution of one thousand recordings per cubic millimeter, the proposed system (time and computing power willing) can record one billion recordings per cubic millimeter. In practice, a resolution of 20 micron was sufficient for non-cellular level FUS characterizations. Successive 2D scans were reliably stacked to form a 3D rendering of the generated acoustic field with the average focal point intensity yielding a 1% coefficient of variation between identical scans. This inherent variation can be used as the threshold of significance for true change detection; to rapidly detect changes to the FUS signal, sampling can be performed at regions of high baseline values. DISCUSSION/SIGNIFICANCE: Focused ultrasound medical devices are gaining popularity for treatments including tumor ablation, neuromodulation, and drug delivery; however, the field lacks a standardized method to characterize these FUS transducers. The presented platform and protocols enable a rigorous and high quality translation through verification and validation.

Perspectives From French and Filipino Parents on the Adaptation of Child Health Knowledge Translation Tools: Qualitative Exploration.

BackgroundA number of evidence-based knowledge translation (KT) tools for parents of children with acute health conditions have been developed. These tools were created and tested with parental input and disseminated to groups proficient in English. Therefore, it is unclear whether they are useful for populations that are more diverse. To enhance the reach of our current and future KT tools, language translation and cultural adaptations may promote relevance for previously underserved knowledge users.ObjectiveThis study aims to explore and understand considerations for the cultural and linguistic adaptation of a KT tool in French and Filipino communities.MethodsA KT tool (whiteboard animation video) describing the signs and symptoms of croup was originally developed in English to provide parents with evidence-based information couched within a narrative reflecting parents’ experiences with the condition. This KT tool was adapted (linguistics and imagery) for French- and Tagalog-speaking parents and caregivers through feedback from key stakeholders. The videos were presented to the respective language speakers for usability testing and discussion. Participants were asked to view the KT tool, complete a usability survey, and participate in semistructured interviews. Audio recordings from the interviews were transcribed verbatim, translated into English, and analyzed for relevant themes by using thematic analysis.ResultsFrench- (n=13) and Tagalog-speaking (n=13) parents completed the usability survey and were interviewed. Although analyzed separately, both data sets produced similar findings, with key themes relating to understanding, relatability, and accessibility. Both the French and Tagalog groups reported that the video and other KT tools were useful in their adapted forms. Participants in both groups cautioned against using verbatim vocabulary and suggested that cultural competency and understanding of health languages were essential for high-quality translations. Parents also discussed their preference for videos with diverse visual representations of families, home environments, and health care workers, as such videos represent their communities more broadly.ConclusionsFrench and Filipino parents appreciated having KT tools in their first language; however, they were also supportive of the use of English KT products. Their suggestions for improving the relatability and communication of health messages are important considerations for the development and adaptation of future KT products. Understanding the needs of the intended end users is a crucial first step in producing relevant tools for health evidence dissemination.

Cultural Memories and Translations of Zhuang Liao Songs

Rooted in ancient Zhuang civilization with unique lyrics and tones, Zhuang Liao Songs have described Zhuang people’s life and work for generations. As the collective cultural memory of the Zhuang minority, Liao Songs have great values in literature, arts, history, folklore, and culture. They were listed in Chinese National Intangible Cultural Heritage in 2008, with their multi-functions of moralizing ideologies, edifying virtues, regulating behavior, and boosting the local economy. The integrative study of memory and its translation has received increasing attention from researchers of intangible cultural heritage. This paper aims to analyze the cultural memory of Zhuang Liao Songs and their intra-lingual and inter-lingual translations for the acceptability and accuracy of English translation in the era of “excellent Chinese culture going out”; to ensure the high-quality translation and dissemination of Zhuang Liao Songs.

Continuous and Diverse Image-to-Image Translation via Signed Attribute Vectors

Recent image-to-image (I2I) translation algorithms focus on learning the mapping from a source to a target domain. However, the continuous translation problem that synthesizes intermediate results between two domains has not been well-studied in the literature. Generating a smooth sequence of intermediate results bridges the gap of two different domains, facilitating the morphing effect across domains. Existing I2I approaches are limited to either intra-domain or deterministic inter-domain continuous translation. In this work, we present an effectively signed attribute vector, which enables continuous translation on diverse mapping paths across various domains. In particular, we introduce a unified attribute space shared by all domains that utilize the sign operation to encode the domain information, thereby allowing the interpolation on attribute vectors of different domains. To enhance the visual quality of continuous translation results, we generate a trajectory between two sign-symmetrical attribute vectors and leverage the domain information of the interpolated results along the trajectory for adversarial training. We evaluate the proposed method on a wide range of I2I translation tasks. Both qualitative and quantitative results demonstrate that the proposed framework generates more high-quality continuous translation results against the state-of-the-art methods.

Innovation and Exploration of Computer-Aided New Media Translation Course Teaching Mode under the Ecological Environment.

In order to analyze the innovation of computer new media-aided translation course teaching mode under the ecological environment, this paper introduces the CorpTrans system, which analyzes the practical effect of computer-aided translation system applied in the course teaching, verifies the role of the system in computer-aided translation course teaching, and plays an important role in improving students' translation ability. At the same time, starting with the development trend of new media environment, the CorpTrans system is used to innovate the teaching mode of computer-aided translation course. The teaching ecology of the comprehensive new media translation curriculum consists of the interaction between the ecological subject of teaching and the ecological environment of teaching. Through the efforts of both teachers and students, high-quality computer-aided translation courses will contribute to cultivating compound translation talents. The innovation of teaching mode needs to be realized by this system, so as to better meet the requirements of improving the teaching level of translation courses.

Automatische vertaling: een levensvatbare oplossing voor het recht?

The translation of legal information constitutes a special branch of translation practice, particularly due to the systemic nature of legal terminology, as well as complicated syntax and the use of fixed word associations and language templates. Professional human translators have always been the preferred way to translate legal documents because they have the linguistic expertise, a good understanding of the cultural context and capacity to make decisions when translating. As more and more international companies enter new markets, the need for translating legal texts increases, which often involves large amounts of data that need to be translated quickly. The possibilities and limits of machine translation (MT) are being frequently tested. Since the legal sector traditionally requires high-quality translations, the question arises whether machine translation can also work well in this field?

Terminology Work in the Translation Project Process. The Hungarian Perspective

    The goal of this paper is to examine the place and role of terminology and terminology work in the translation project process and, based on literature and everyday practice, to examine the changes relevant to terminology work that have taken place in Hungary in recent years. The method of analysis is a comparative content analysis and descriptive method.    Terminology management, document management, and translation (project) management are all interrelated as are all forms of content management, and terminology has a key role in all of them. The quality of the source text is fundamental for the quality of the translation, therefore high-quality translation begins at content creation. Starting with drafting the documentation, the translation project process is a fourteen-step project process. Terminology work is present in most steps of the translation project process in some way, and even plays an important role in step zero, the creation of the source text. Terminology work, both in the translation project process and in the writing of documentation, has a number of regulated processes and steps, some of which are laid down in standards and are also discussed in the international literature on terminology, translation, and technical communication. In daily practice, translation agencies are increasingly commissioned to carry out documentation work in addition to translation, editing, and proofreading tasks. Nevertheless, it was found that in Hungary, in the operation of translation agencies and during the process of technical documentation, terminology workflows are not always coordinated, terminology databases are not always used, and terminology specialists are rarely employed.

Діалектна основа перекладу Михайла Атаманюка німецькомовного роману К.Е. Францоза «Боротьба за право»

The article focuses on the first Ukrainian translation of the novel Ein Kampf um’s Recht by the Austrian writer Karl Emil Francoz of the mid-19th century, which tells about the struggle of Galician peasants - Rusins - for their land. The translation was made in Chernivtsi in 1910 by Mykhailo Atamanyuk, a Ukrainian writer and translator, whose creative legacy has been unfairly overlooked by researchers. Mykhailo Atamanyuk grew up and got education in Austro-Hungary, where he witnessed the interethnic dialogue and intercuLturaL contacts and mastered the local dialect from his childhood. These factors, together with Atamanyuk's consonance with Franzos in the perpception of reality, formed a reliable basis for a high-quality translation into the native language of its heroes. The authors argue that the translation under analysis accurately conveys all the typical features of the local dialect. The article analyses the translation in terms of the dialectal facts on different language levels and identifies phonetic, morphological, lexical, and syntactic dialectal phenomena, typical for the Transnistrian dialect of the Galician-Bukovinian dialects within the South-Western dialect of the Ukrainian language. Basing on the two toponyms mentioned in the novel, the authors determine its linguo-spatial background - Kolomyisky, Tlumatsky, Horodenkivsky, and Snyatinsky districts of Ivano-Frankivsk Region of modern Ukraine, which represent the Transnistrian dialect at the Ukrainian dialect map. These dialects formed the basis of the Western Ukrainian variety of the literary language, which contributed to the popularity of the novel among Ukrainian readers. Thus, for the first time, the dialect basis of Atamanyuk's Ukrainian translation of the novel Ein Kampf um’s Recht by K.E. Franzos has been established.

The main problems of translation of the novel “rebecca” by daphne du maurier from English into Uzbek

This article deals with the main problems of the translation of the novel “Rebecca” by Daphne Du Maurier from English into Uzbek with providing some kind of recommendations to a high-quality translation of the literature. The main points of the article are about the process of translation of English literature, translation studies, stylistics and its analyses. Expressive means and stylistic devices as well as their use in Uzbek and English literature are shown by examples in Daphne Du Maurier's work. During translating the national words which express the colorfulness of the country, phrases, sayings, and proverbs are paid attention and translated by finding the appropriate form in another language. Besides, two languages were compared while analyzing the literary works.

Experts, Errors, and Context: A Large-Scale Study of Human Evaluation for Machine Translation

AbstractHuman evaluation of modern high-quality machine translation systems is a difficult problem, and there is increasing evidence that inadequate evaluation procedures can lead to erroneous conclusions. While there has been considerable research on human evaluation, the field still lacks a commonly accepted standard procedure. As a step toward this goal, we propose an evaluation methodology grounded in explicit error analysis, based on the Multidimensional Quality Metrics (MQM) framework. We carry out the largest MQM research study to date, scoring the outputs of top systems from the WMT 2020 shared task in two language pairs using annotations provided by professional translators with access to full document context. We analyze the resulting data extensively, finding among other results a substantially different ranking of evaluated systems from the one established by the WMT crowd workers, exhibiting a clear preference for human over machine output. Surprisingly, we also find that automatic metrics based on pre-trained embeddings can outperform human crowd workers. We make our corpus publicly available for further research.

‘Turning the Wheel of the Teaching’: A translation of Aśvaghoṣa’s Buddhacarita Canto 15 from a recently rediscovered Sanskrit manuscript

Asian Literature and Translation offers high quality translations of Asian literary sources and reflections on the theoretical and practical issues that attend upon translation (both in the past and present). It also publishes articles and comment pieces on Asian textual sources more generally, as well as work on the reception of Asian literature within and beyond the region.

On Mean Dependency Distance as a Metric of Translation Quality Assessment

This paper has adopted a quantitative approach to carry out a linguistic study, within the theoretical framework of dependency grammar. Translation is a process where source language and target language interact with each other. The present study aims at exploring the feasibility of mean dependency distance as a metric for automated translation quality assessment. The current research hypothesized that different levels of translation are significantly different in the aspect of mean dependency distance. Data of this study were based on the written translation in Parallel Corpus of Chinese EFL Learners which was composed of translations from Chinese EFL learners in various topic. The translations were human-scored to determine the levels of translation, according to which the translations were categorized. Our results indicated that: (1) senior students perform better in translation than junior students, and mean dependency distance of translations from senior group is significantly shorter than the junior; (2) high quality translations yield shorter mean dependency distance than the low quality translations; (3) mean dependency distance of translations is moderately correlated with the human score. The resultant implication suggests the potential for mean dependency distance in differentiating translations of different quality.