To see the other types of publications on this topic, follow the link: Data translation model.
Journal articles on the topic 'Data translation model'
Create a spot-on reference in APA, MLA, Chicago, Harvard, and other styles
Consult the top 50 journal articles for your research on the topic 'Data translation model.'
Next to every source in the list of references, there is an 'Add to bibliography' button. Press on it, and we will generate automatically the bibliographic reference to the chosen work in the citation style you need: APA, MLA, Harvard, Chicago, Vancouver, etc.
You can also download the full text of the academic publication as pdf and read online its abstract whenever available in the metadata.
Browse journal articles on a wide variety of disciplines and organise your bibliography correctly.
1
Marie, Benjamin, and Atsushi Fujita. "Synthesizing Parallel Data of User-Generated Texts with Zero-Shot Neural Machine Translation." Transactions of the Association for Computational Linguistics 8 (November 2020): 710–25. http://dx.doi.org/10.1162/tacl_a_00341.
Full textAbstract:
Neural machine translation (NMT) systems are usually trained on clean parallel data. They can perform very well for translating clean in-domain texts. However, as demonstrated by previous work, the translation quality significantly worsens when translating noisy texts, such as user-generated texts (UGT) from online social media. Given the lack of parallel data of UGT that can be used to train or adapt NMT systems, we synthesize parallel data of UGT, exploiting monolingual data of UGT through crosslingual language model pre-training and zero-shot NMT systems. This paper presents two different but complementary approaches: One alters given clean parallel data into UGT-like parallel data whereas the other generates translations from monolingual data of UGT. On the MTNT translation tasks, we show that our synthesized parallel data can lead to better NMT systems for UGT while making them more robust in translating texts from various domains and styles.
2
Li, Rumeng, Xun Wang, and Hong Yu. "MetaMT, a Meta Learning Method Leveraging Multiple Domain Data for Low Resource Machine Translation." Proceedings of the AAAI Conference on Artificial Intelligence 34, no. 05 (April 3, 2020): 8245–52. http://dx.doi.org/10.1609/aaai.v34i05.6339.
Full textAbstract:
Neural machine translation (NMT) models have achieved state-of-the-art translation quality with a large quantity of parallel corpora available. However, their performance suffers significantly when it comes to domain-specific translations, in which training data are usually scarce. In this paper, we present a novel NMT model with a new word embedding transition technique for fast domain adaption. We propose to split parameters in the model into two groups: model parameters and meta parameters. The former are used to model the translation while the latter are used to adjust the representational space to generalize the model to different domains. We mimic the domain adaptation of the machine translation model to low-resource domains using multiple translation tasks on different domains. A new training strategy based on meta-learning is developed along with the proposed model to update the model parameters and meta parameters alternately. Experiments on datasets of different domains showed substantial improvements of NMT performances on a limited amount of data.
3
Cao, Qianyu, and Hanmei Hao. "A Chaotic Neural Network Model for English Machine Translation Based on Big Data Analysis." Computational Intelligence and Neuroscience 2021 (July 2, 2021): 1–12. http://dx.doi.org/10.1155/2021/3274326.
Full textAbstract:
In this paper, the chaotic neural network model of big data analysis is used to conduct in-depth analysis and research on the English translation. Firstly, under the guidance of the translation strategy of text type theory, the translation generated by the machine translation system is edited after translation, and then professionals specializing in computer and translation are invited to confirm the translation. After that, the errors in the translations generated by the machine translation system are classified based on the Double Quantum Filter-Muttahida Quami Movement (DQF-MQM) error type classification framework. Due to the characteristics of the source text as an informative academic text, long and difficult sentences, passive voice, and terminology translation are the main causes of machine translation errors. In view of the rigorous logic of the source text and the fixed language steps, this research proposes corresponding post-translation editing strategies for each type of error. It is suggested that translators should maintain the logic of the source text by converting implicit connections into explicit connections, maintain the academic accuracy of the source text by adding subjects and adjusting the word order to deal with the passive voice, and deal with semitechnical terms by appropriately selecting word meanings in postediting. The errors of machine translation in computer science and technology text abstracts are systematically categorized, and the corresponding post-translation editing strategies are proposed to provide reference suggestions for translators in this field, to improve the quality of machine translation in this field.
4
Yu, Lei, Laurent Sartran, Wojciech Stokowiec, Wang Ling, Lingpeng Kong, Phil Blunsom, and Chris Dyer. "Better Document-Level Machine Translation with Bayes’ Rule." Transactions of the Association for Computational Linguistics 8 (July 2020): 346–60. http://dx.doi.org/10.1162/tacl_a_00319.
Full textAbstract:
We show that Bayes’ rule provides an effective mechanism for creating document translation models that can be learned from only parallel sentences and monolingual documents a compelling benefit because parallel documents are not always available. In our formulation, the posterior probability of a candidate translation is the product of the unconditional (prior) probability of the candidate output document and the “reverse translation probability” of translating the candidate output back into the source language. Our proposed model uses a powerful autoregressive language model as the prior on target language documents, but it assumes that each sentence is translated independently from the target to the source language. Crucially, at test time, when a source document is observed, the document language model prior induces dependencies between the translations of the source sentences in the posterior. The model’s independence assumption not only enables efficient use of available data, but it additionally admits a practical left-to-right beam-search algorithm for carrying out inference. Experiments show that our model benefits from using cross-sentence context in the language model, and it outperforms existing document translation approaches.
5
Luo, Gong-Xu, Ya-Ting Yang, Rui Dong, Yan-Hong Chen, and Wen-Bo Zhang. "A Joint Back-Translation and Transfer Learning Method for Low-Resource Neural Machine Translation." Mathematical Problems in Engineering 2020 (May 31, 2020): 1–11. http://dx.doi.org/10.1155/2020/6140153.
Full textAbstract:
Neural machine translation (NMT) for low-resource languages has drawn great attention in recent years. In this paper, we propose a joint back-translation and transfer learning method for low-resource languages. It is widely recognized that data augmentation methods and transfer learning methods are both straight forward and effective ways for low-resource problems. However, existing methods, which utilize one of these methods alone, limit the capacity of NMT models for low-resource problems. In order to make full use of the advantages of existing methods and further improve the translation performance of low-resource languages, we propose a new method to perfectly integrate the back-translation method with mainstream transfer learning architectures, which can not only initialize the NMT model by transferring parameters of the pretrained models, but also generate synthetic parallel data by translating large-scale monolingual data of the target side to boost the fluency of translations. We conduct experiments to explore the effectiveness of the joint method by incorporating back-translation into the parent-child and the hierarchical transfer learning architecture. In addition, different preprocessing and training methods are explored to get better performance. Experimental results on Uygur-Chinese and Turkish-English translation demonstrate the superiority of the proposed method over the baselines that use single methods.
6
Wahiyudin, Ummi Nadjwa, and Taj Rijal Bin Muhamad Romli. "Tanslating Malay Compounds into Arabic Based on Dynamic Theory and Arabization Method." Journal of Islamic Thought and Civilization 11, no. 1 (June 28, 2021): 43–58. http://dx.doi.org/10.32350/jitc.111.03.
Full textAbstract:
This study aims at making possible the effective use of machine translation (MT) in interpreting the Malay compounds into Arabic ones following the structure and Arabic style. The necessity of this study arises on account of the weakness of translation quality using online MT and the lack of suitable methods to structure the compounds from the Malay language into Arabic. There are three objectives of this study which are to collect the results of Malay compound translations using online MT into Arabic, analyze the results of the compound translations, and suggest compound translation methods based on dynamic theory and Arabization method. The study uses three online MT as instruments to translate: Google Translate, Microsoft Bing Translator, and Yandex Translator. This qualitative study employs a descriptive approach and analysis method in collecting information and analyzing data. The study focuses on 15 Malay compounds which are later categorized into school names, hospital names, and clinics. The findings of translation have been drawn using the next three MTs and analyzed at three main level: namely grammar level, phonetics and phonology level, and dynamic translation level. From this analysis, 4 out of 15 compound nouns translations data into Arabic are categorized as poor translations for not approaching the structure and Arabic style. In the final stages, the results of the translation collected are formulated and suggested alternative translations based on dynamic theory and methods of Arabization and compound restructuring formula in Arabic. Through this process, the translation results of the compounds can be categorized as translations that can meet the structure and style of the Arabic language. The compound translation model can be proposed as a new translation method for Arabic language users, especially the Arabic translators and students both at school and higher education.
7
Robin, Edina. "Translation universals revisited." FORUM / Revue internationale d’interprétation et de traduction / International Journal of Interpretation and Translation 15, no. 1 (August 19, 2017): 51–66. http://dx.doi.org/10.1075/forum.15.1.03rob.
Full textAbstract:
Abstract According to the results of translation-based empirical research within the descriptive paradigm, transfer operations and the shifts that occur as a result of translators’ interventions are governed by norms, which represent general, standard practices built on informal social consensus (Toury 1995). Based on the scientific analysis of norms and general rules, the so-called translation universals were formulated describing the factors and qualities that distinguish translations from source texts and from authentic texts not produced through translation but originally written in the target language (Baker 1993). In the present study, I aim to summarise the theoretical conclusions drawn so far from the description of these observed translational features, as well as the results of the research into linguistic phenomena and laws that characterise translations in general, then I will synthesise and graphically represent the lessons learned in a theoretical model. Hopefully, it will provide help to understand and process the research data gained so far and in the future.
8
Jassem, Krzysztof, and Tomasz Dwojak. "Statistical versus neural machine translation – a case study for a medium size domain-specific bilingual corpus." Poznan Studies in Contemporary Linguistics 55, no. 2 (June 26, 2019): 491–515. http://dx.doi.org/10.1515/psicl-2019-0018.
Full textAbstract:
Abstract Neural Machine Translation (NMT) has recently achieved promising results for a number of translation pairs. Although the method requires larger volumes of data and more computational power than Statistical Machine Translation (SMT), it is believed to become dominant in near future. In this paper we evaluate SMT and NMT models learned on a domain-specific English-Polish corpus of a moderate size (1,200,000 segments). The experiment shows that both solutions significantly outperform a general-domain online translator. The SMT model achieves a slightly better BLEU score than the NMT model. On the other hand, the process of decoding is noticeably faster in NMT. Human evaluation carried out on a sizeable sample of translations (2,000 pairs) reveals the superiority of the NMT approach, particularly in the aspect of output fluency.
9
Faria Shaheen, Dr. Ghulam Ali, and Dr. Kanwal Zahra. "Translating Feminist Identities: A Critical Discourse Analysis of Urdu Translation of Brown’s Work ‘The Dancing Girls of Lahore’." Research Journal of Social Sciences and Economics Review (RJSSER) 2, no. 1 (March 16, 2021): 330–37. http://dx.doi.org/10.36902/rjsser-vol2-iss1-2021(330-337).
Full textAbstract:
The present paper focuses on the construction and production of gender identities through discursive mechanisms during the process of translation. It also attempts to focus on the various discursive strategies used by the translator in the phenomenon of translation. The data comprise Louis Brown's book 'The Dancing Girls of Lahore' and its Urdu translation by Pakistani male writer and translator Dr. Naeem Tariq. The theoretical and conceptual framework for the present research is based on three dimensional model by Farahzad (2012) in the light of Critical Discourse Analysis. The data related to feminist discourse in both the source text and the target text is selected through the purposive sampling technique. The analysis of the present paper reflects the position of the translator through his lexical and grammatical choices in the process of translation. The findings of the present study reflect the position of male translators and expose the socio-political structure of Pakistani patriarchal society. This study provides various dimensions to explore feminist translations for future researchers.
10
Islam Saleem, Haneen, and Hakar Hazim M.Ameen. "Problems Students Face in Translating Discourse Markers from English into Kurdish." Academic Journal of Nawroz University 10, no. 3 (July 13, 2021): 1–8. http://dx.doi.org/10.25007/ajnu.v10n3a1067.
Full textAbstract:
This paper investigates the problems are found in the translation of discourse markers from English into Kurdish. One of the main problems that discourse markers cause for translators is that discourse markers may have various possible translation choices. Moreover, a discourse marker has many functions, i.e. it may have more than one function. It can be thus used to show a variety of relations between several written discourse parts. Accordingly, a translator has to decide the function that a discourse marker has in a definite context to give the correct translation. In this paper Fraser's Model of discourse markers analysis (1999) has been chosen for investigating discourse markers. As for the data analysis, four students of the Department of Translation at the University of Duhok have been chosen to translate the same text from English into Kurdish to show the difficulties in choosing the correct discourse markers in these four different translations. The results showed that there are no equivalents in choosing the correct discourse markers when translating them into Kurdish.
11
Alhassan, Awad, Yasser Muhammad Naguib Sabtan, and Lamis Omar. "Using Parallel Corpora in the Translation Classroom: Moving towards a Corpus-driven Pedagogy for Omani Translation Major Students." Arab World English Journal 12, no. 1 (March 15, 2021): 40–58. http://dx.doi.org/10.24093/awej/vol12no1.4.
Full textAbstract:
Research has shown that parallel corpora have potential benefits for translator training and education. Most of the current available Arabic corpora, modern standard or dialectical, are monolingual in nature and there is an apparent lack in the Arabic-English parallel corpora for translation classroom. The present study was aimed to investigate the translation problems encountered by Omani translation major students when translating from Arabic into English with a view to proposing some corpus-informed pedagogy approach for training student translators to overcome these challenges by looking at some model samples of professional translation. Thirty students voluntarily took part in the investigation. The study adopted a combination of both corpus and qualitative methodology whereby some typical problems students would encounter when translating from Arabic-into-English were selected along with some specific Arabic texts involving these problems were prepared and the participants were asked to translate them into English. The participants were provided with some samples of the parallel English translated texts and were asked to compare and contrast their translations with these samples and reflect on the overall experience. They were then interviewed to explore their impressions about and the extent to which they think that parallel corpora would help them improve their translation. Results of data analysis indicated that the participants experienced several translation challenges. They, however, showed an overall positive attitude towards parallel Arabic-English corpora as they reportedly found them very helpful in improving their translation. Pedagogical implications for corpus-informed translation teaching, training and materials design and development are presented and discussed.
12
Zhang, Wenbo, Xiao Li, Yating Yang, and Rui Dong. "Pre-Training on Mixed Data for Low-Resource Neural Machine Translation." Information 12, no. 3 (March 18, 2021): 133. http://dx.doi.org/10.3390/info12030133.
Full textAbstract:
The pre-training fine-tuning mode has been shown to be effective for low resource neural machine translation. In this mode, pre-training models trained on monolingual data are used to initiate translation models to transfer knowledge from monolingual data into translation models. In recent years, pre-training models usually take sentences with randomly masked words as input, and are trained by predicting these masked words based on unmasked words. In this paper, we propose a new pre-training method that still predicts masked words, but randomly replaces some of the unmasked words in the input with their translation words in another language. The translation words are from bilingual data, so that the data for pre-training contains both monolingual data and bilingual data. We conduct experiments on Uyghur-Chinese corpus to evaluate our method. The experimental results show that our method can make the pre-training model have a better generalization ability and help the translation model to achieve better performance. Through a word translation task, we also demonstrate that our method enables the embedding of the translation model to acquire more alignment knowledge.
13
Atzeni, Paolo, Giorgio Gianforme, and Paolo Cappellari. "Data model descriptions and translation signatures in a multi-model framework." Annals of Mathematics and Artificial Intelligence 63, no. 3-4 (December 2011): 287–315. http://dx.doi.org/10.1007/s10472-012-9277-y.
Full text
14
Mohammadpour, Fahime, Mohammadtaghi Shahnazari-Dorcheh, and Mahmoud Afrouz. "Looking through the lens of Bourdieu: A corpus-based Study of English Romance Fiction Translation." Hikma 19, no. 2 (December 22, 2020): 327–55. http://dx.doi.org/10.21071/hikma.v19i2.12871.
Full textAbstract:
Habitus is one of the key concepts of the Bourdieusian sociology which Translation Studies has benefited. Based on the Bourdieusian sociological model, this study investigated the translatorial habitus of the Iranian translators of English romance novels as far as the translation strategies of culture-specific items (CSIs) are concerned before and after the Cultural Revolution of 1980 in Iran. The research data include 4282 sentences containing CSIs extracted from Rebecca, Sense and Sensibility, and The Great Gatsby, and their two Persian translations. The extracted data were analyzed, adopting a consolidated typology of translation procedures for CSIs. The strategies employed for translating CSIs are presented with frequencies and percentages using descriptive statistics. Moreover, the results were corroborated with a qualitative analysis of some archived interviews printed in Motarjem [the translator] journal. The investigation revealed three essential findings: a marked source-oriented tendency among Iranian translators of the English romance novels when translating CSIs in the Pre-Cultural Revolution era, maintaining the same tendency in the Post-Cultural Revolution era, and finally a growing tendency in moving from Pre- to Post-Cultural Revolution era. The results of the Chi-square test highlighted a significant difference between various strategies used in two eras.
15
Meiliana, Sylvie. "The implementation of literary works In teaching literary translation." EDUTEC : Journal of Education And Technology 4, no. 1 (September 30, 2020): 28–40. http://dx.doi.org/10.29062/edu.v4i1.81.
Full textAbstract:
The aim of this research is to show the implementation of literary works in literary translation by giving the way how to implement the literary work in revealing cultural terms found in a literary work, namely Achmad Tohari’s Ronggeng Dukuh Paruk Novel. Based on the literary work, the research on literary translation is done by investigating the translation procedure applied in translating the cultural terms from Indonesian into English. This research used a descriptive qualitative method with content analysis technique done by taking the flow model followed by data reduction, data presentation, and conclusion. The analysis used a semantic approach and Peter Newmark’s translation procedures. Result of the research shown by implementing literary work, the research of literary translation reveals that there are 16 cultural terms and classified in 6 different categories, they are musical instruments, clothes, accessories, work and leisure, activities and procedures, and religious terms. In translating the novel, there are 7 translation procedures used by the translator, they are transference, naturalization, cultural equivalent, functional equivalent, descriptive equivalent, couplets, and notes.
16
Ardianto, Ardik. "Translation of Addressing Terms in The Novel This Earth of Mankind." Deskripsi Bahasa 2, no. 2 (October 7, 2019): 119–27. http://dx.doi.org/10.22146/db.v2i2.351.
Full textAbstract:
This paper is an attempt to approach the translational stylistics, aiming at identifying the equivalence and translation procedures used in translating the Toer’s authorial style from the Indonesian language to the English language in the novel This Earth of Mankind. A translational stylistics model proposed by Malmkjær was used to contrast the target text (TT) and the source text (ST), primarily focusing on the stylistic shift. Further, as to the model of translation procedures, it specifically employed Vinay and Darbelnet’s methodology for translation. Data used in this study were addressing terms found in two novels, the Indonesian novel Bumi Manusia and its translation This Earth of Mankind. The rigorous analysis demonstrated how the translation of addressing terms involved a wide range of aspects, such as sociocultural and historical values (including social identity and social strata) and power and solidarity relation. Therefore, it raised a number of noteworthy translation issues, i.e., its equivalence, stylistic shift, and translator’s strategies. Through the increasing awareness of ‘cultural turn’ in translation studies, the concept of equivalence is supposedly perceived not as an absolute assessment but as a mediating attempt to accommodate and transpose the inferred or perceived meaning from the ST to the TT as much as possible. However, the findings are not set out to appraise the translator’s ethical attitude, considering the limited data used in this study and numerous factors that are not yet taken into account, e.g. the power play of the translation industry, and culture-mediating agenda in the receiving culture.
17
Hasyim, Muhammad, Prasuri Kuswarini, and Kaharuddin. "- SEMIOTIC MODEL FOR EQUIVALENCE AND NON-EQUIVALENCE IN TRANSLATION." Humanities & Social Sciences Reviews 8, no. 3 (May 20, 2020): 381–91. http://dx.doi.org/10.18510/hssr.2020.8341.
Full textAbstract:
Purpose of the study: Not all languages have a universal concept of the same object, and this creates problems in translation. This paper aims to examine the semiotic model for equivalence or non-equivalence in translation which attempts to define the semiotic model, to use the model for translation, and to offer the benefits of this model to solving translation’s problem in equivalence and non-equivalence.
Methodology: The data of this research are derived from the novel Lelaki Harimau, as the source language and L'homme Tigre, as the target language. This model is used in the Indonesian novel which has been translated into 14 languages, one of which is in French. The authors use a semiotic approach to analyze the equivalence and non-equivalence in the translation.
Main Findings: This study reveals that the concept of signified in the semiotic theory proposes two models: the first: translation using the same concept in the source text (ST) and target text (TT), which is broadly known as equivalence, the second: translation using different concept between ST and TT, this called non-equivalence. This article not only explores the issue of meaning contextually in translation, but also the use of the semiotic model in translation which shows that the language perspective depends on the relationship between the sign and the object.
Applications of this study: The model for this study can be used not only in translation studies at universities but also in providing supporting data for applied linguistic studies.
Novelty/Originality of this study: This study provides a novelty in translation research with a semiotic approach. The contribution of this study is that the semiotics perspective suggests that a sign in the concept level (signified) will not be universal due to different cultural backgrounds.
18
Mohamed, Entesar Alsir Abu-ALgasim. "Semantic Problems of the Usage of Archaic Morphological Features: Surat Al-Humza (Traducer) as a Model." International Journal of Linguistics, Literature and Translation 4, no. 2 (February 27, 2021): 138–45. http://dx.doi.org/10.32996/ijllt.2021.4.2.17.
Full textAbstract:
This study presents a comparative analysis of the translation of the holy Qur'anic verses by Pickthall and Yusuf Ali from one hand, and the translation of Sahih International from the other. By tracing the first translation style, it had been found that they tend to use archaism or old words to a large extent as they trying to make their translation sound like 'scripture' to an English- speaking audience, and this results in loss of meaning pose semantic difficulty in translating the Holy verses. Abdalati M. Ali in his paper (Lexical and Semantic Problems in Translating Quran) proves that Pickthall and Yusuf Ali tend to use archaism in their translation, as he puts that ‘'the translations of Pickthall and Ali appear to be written in an archaic form of English''. In contrast to the first style the Sahih International translation aims to provide a literal rendering of the Arabic of the Qur’an into “plain” English. It has become popular as a more contemporary translation, but focuses more on providing a literal meaning of the Arabic than on providing smooth English. This makes it a useful resource for students who are seeking to learn the literal meaning of the Arabic of the Qur’an. Being a text at the highest degree of eloquence, the Qur’ān, as a whole, poses a serious challenge for translators and linguists alike. The challenging areas within the Qur'anic text are indeed too numerous to count. This paper investigates one of the major problem areas when translating the Qur’ān, namely, archaic words. This linguistic feature is intrinsic to the Qur'anic text and, in turn, has an impact on its translation leading to ambiguity. In this regards the rendition of qur'anic verses is considered as a difficult mission, this is due to Quran is written in a highly symbolic and classical form of the Arabic language, therefore, rendering it requires a deep knowledge and grasping of its meanings in addition to that translators should be able to reflect those meanings into the target language. Based on the mentioned facts usage of archaic and old words will constitute obstacle because it increases the complexities of target text, thus, it will distort the implicit meaning, and confuse target readers. Analytical descriptive method of data collection has been followed which comprises tools, samples, procedures.
19
Chabiniok, Radomir, Vicky Y. Wang, Myrianthi Hadjicharalambous, Liya Asner, Jack Lee, Maxime Sermesant, Ellen Kuhl, et al. "Multiphysics and multiscale modelling, data–model fusion and integration of organ physiology in the clinic: ventricular cardiac mechanics." Interface Focus 6, no. 2 (April 6, 2016): 20150083. http://dx.doi.org/10.1098/rsfs.2015.0083.
Full textAbstract:
With heart and cardiovascular diseases continually challenging healthcare systems worldwide, translating basic research on cardiac (patho)physiology into clinical care is essential. Exacerbating this already extensive challenge is the complexity of the heart, relying on its hierarchical structure and function to maintain cardiovascular flow. Computational modelling has been proposed and actively pursued as a tool for accelerating research and translation. Allowing exploration of the relationships between physics, multiscale mechanisms and function, computational modelling provides a platform for improving our understanding of the heart. Further integration of experimental and clinical data through data assimilation and parameter estimation techniques is bringing computational models closer to use in routine clinical practice. This article reviews developments in computational cardiac modelling and how their integration with medical imaging data is providing new pathways for translational cardiac modelling.
20
Ilani, Ali, and Hossein Barati. "Translations of Journalistic Texts in Iranian Undergraduate Students: An Error Analysis Approach." International Journal of English Linguistics 6, no. 6 (November 24, 2016): 147. http://dx.doi.org/10.5539/ijel.v6n6p147.
Full textAbstract:
<p>Translating journalistic text has been one of the major courses in Iranian universities. The challenges hidden in translating journalistic texts motivated the present study to investigate the translation of such texts. Thus, this research makes an attempt to identify and categorize the probable errors and to distinguish the most frequent ones. Furthermore, it tries to find whether there is a pattern among the errors committed by students in their translations. To this end, a translation test of Persian journalistic texts was developed. Forty students studying English translation were recruited for this study. In order to analyze collected data, Keshavarz’s Model (1997) and ATA were used for error analysis. The current study found that there is not a pattern among errors committed by students. The most frequent errors were categorized as (i) grammar, (ii) terminology, and (iii) misunderstanding of original text.</p>
21
Poirier, Éric. "Exploring theoretical functions of corpus data in teaching translation." Cadernos de Tradução 36, no. 1 (April 26, 2016): 177. http://dx.doi.org/10.5007/2175-7968.2016v36nesp1p177.
Full textAbstract:
http://dx.doi.org/10.5007/2175-7968.2016v36nesp1p177As language referential data banks, corpora are instrumental in the exploration of translation solutions in bilingual parallel texts or conventional usages of source or target language in monolingual general or specialized texts. These roles are firmly rooted in translation processes, from analysis and interpretation of source text to searching for an acceptable equivalent and integrating it into the production of the target text. Provided the creative and not the conservative way be taken, validation or adaptation of target text in accordance with conventional usages in the target language also benefits from corpora. Translation teaching is not exploiting this way of translating that is common practice in the professional translation markets around the world. Instead of showing what corpus tools can do to translation teaching, we start our analysis with a common issue within translation teaching and show how corpus data can help to resolve it in learning activities in translation courses. We suggest a corpus-driven model for the interpretation of ‘business’ as a term and as an item in complex terms based on source text pattern analysis. This methodology will make it possible for teachers to explain and justify interpretation rules that have been defined theoretically from corpus data. It will also help teachers to conceive and non-subjectively assess practical activities designed for learners of translation. Corpus data selected for the examples of rule-based interpretations provided in this paper have been compiled in a corpus-driven study (Poirier, 2015) on the translation of the noun ‘business’ in the field of specialized translation in business, economics, and finance from English to French. The corpus methodology and rule-based interpretation of senses can be generalized and applied in the definition of interpretation rules for other language pairs and other specialized simple and complex terms. These works will encourage the matching of translation study theories and corpus translation studies with professional practices. It will also encourage the matching of translation studies and corpus translation studies with source and target language usages and with textual correlations between source language real usages and target language translation real practices.
22
Sun, Zewei, Shujian Huang, Hao-Ran Wei, Xin-yu Dai, and Jiajun Chen. "Generating Diverse Translation by Manipulating Multi-Head Attention." Proceedings of the AAAI Conference on Artificial Intelligence 34, no. 05 (April 3, 2020): 8976–83. http://dx.doi.org/10.1609/aaai.v34i05.6429.
Full textAbstract:
Transformer model (Vaswani et al. 2017) has been widely used in machine translation tasks and obtained state-of-the-art results. In this paper, we report an interesting phenomenon in its encoder-decoder multi-head attention: different attention heads of the final decoder layer align to different word translation candidates. We empirically verify this discovery and propose a method to generate diverse translations by manipulating heads. Furthermore, we make use of these diverse translations with the back-translation technique for better data augmentation. Experiment results show that our method generates diverse translations without a severe drop in translation quality. Experiments also show that back-translation with these diverse translations could bring a significant improvement in performance on translation tasks. An auxiliary experiment of conversation response generation task proves the effect of diversity as well.
23
Sperber, Matthias, Graham Neubig, Jan Niehues, and Alex Waibel. "Attention-Passing Models for Robust and Data-Efficient End-to-End Speech Translation." Transactions of the Association for Computational Linguistics 7 (November 2019): 313–25. http://dx.doi.org/10.1162/tacl_a_00270.
Full textAbstract:
Speech translation has traditionally been approached through cascaded models consisting of a speech recognizer trained on a corpus of transcribed speech, and a machine translation system trained on parallel texts. Several recent works have shown the feasibility of collapsing the cascade into a single, direct model that can be trained in an end-to-end fashion on a corpus of translated speech. However, experiments are inconclusive on whether the cascade or the direct model is stronger, and have only been conducted under the unrealistic assumption that both are trained on equal amounts of data, ignoring other available speech recognition and machine translation corpora. In this paper, we demonstrate that direct speech translation models require more data to perform well than cascaded models, and although they allow including auxiliary data through multi-task training, they are poor at exploiting such data, putting them at a severe disadvantage. As a remedy, we propose the use of end- to-end trainable models with two attention mechanisms, the first establishing source speech to source text alignments, the second modeling source to target text alignment. We show that such models naturally decompose into multi-task–trainable recognition and translation tasks and propose an attention-passing technique that alleviates error propagation issues in a previous formulation of a model with two attention stages. Our proposed model outperforms all examined baselines and is able to exploit auxiliary training data much more effectively than direct attentional models.
24
Carrasco, Ramón A., Francisco Muñoz-Leiva, and Miguel J. Hornos. "A multidimensional data model using the fuzzy model based on the semantic translation." Information Systems Frontiers 15, no. 3 (January 27, 2013): 351–70. http://dx.doi.org/10.1007/s10796-012-9398-1.
Full text
25
Nabende, Peter. "A Review and evaluation of Machine Translation methods for Lumasaaba." Journal of Digital Science, no. 1 (May 28, 2020): 3–17. http://dx.doi.org/10.33847/2686-8296.2.1_1.
Full textAbstract:
Natural Language Processing for under-resourced languages is now a mainstream research area. However, there are limited studies on Natural Language Processing applications for many indigenous East African languages. As a contribution to covering the current gap of knowledge, this paper focuses on evaluating the application of well-established machine translation methods for one heavily under-resourced indigenous East African language called Lumasaaba. Specifically, we review the most common machine translation methods in the context of Lumasaaba including both rule-based and data-driven methods. Then we apply a state of the art data-driven machine translation method to learn models for automating translation between Lumasaaba and English using a very limited data set of parallel sentences. Automatic evaluation results show that a transformer-based Neural Machine Translation model architecture leads to consistently better BLEU scores than the recurrent neural network-based models. Moreover, the automatically generated translations can be comprehended to a reasonable extent and are usually associated with the source language input.
26
Park, Ocksue. "What constitutes a translation graduate school in South Korea?" Babel. Revue internationale de la traduction / International Journal of Translation 53, no. 2 (November 29, 2007): 167–78. http://dx.doi.org/10.1075/babel.53.2.07par.
Full textAbstract:
South Korea is the first country where the translation education has operated at a postgraduate level in Asia. The first graduate school, the Graduate School of Interpretation and Translation, Hankuk University of Foreign Studies, operated at 1979. As of 2006, there are ten translation and/or interpreting departments at graduate level in South Korea. The curriculum of translation graduate schools in South Korea should be examined from a theoretical standpoint. This article is the answer of what constitutes a translation graduate school in South Korea in terms of its title, its commencement date, type of school, division of departments, admission requirements, and the duration of the programme and what is an appropriate curriculum for a translation graduate school in South Korea. For this study, I have conducted interviews with heads of translation departments at South Korea’s graduate schools in the first place, with a view to establish the real situation of graduate schools that teach translation in South Korea. I adopted Renfer’s model for analysing the programme of the graduate schools. Renfer (1991) presents four basic programme models of translator and/or interpreter education training for western countries such as Two-tier system, Parallel translator and interpreter training model, Y model, and Postgraduate interpreter training or intensive on-the-job training in international organisations. The results of the data analysis are discussed, and lastly suggestions for translation students to develop and improve their translation skills are presented.
27
Huang, Jin-Xia, Kyung-Soon Lee, and Young-Kil Kim. "Hybrid Translation with Classification: Revisiting Rule-Based and Neural Machine Translation." Electronics 9, no. 2 (January 21, 2020): 201. http://dx.doi.org/10.3390/electronics9020201.
Full textAbstract:
This paper proposes a hybrid machine-translation system that combines neural machine translation with well-developed rule-based machine translation to utilize the stability of the latter to compensate for the inadequacy of neural machine translation in rare-resource domains. A classifier is introduced to predict which translation from the two systems is more reliable. We explore a set of features that reflect the reliability of translation and its process, and training data is automatically expanded with a small, human-labeled dataset to solve the insufficient-data problem. A series of experiments shows that the hybrid system’s translation accuracy is improved, especially in out-of-domain translations, and classification accuracy is greatly improved when using the proposed features and the automatically constructed training set. A comparison between feature- and text-based classification is also performed, and the results show that the feature-based model achieves better classification accuracy, even when compared to neural network text classifiers.
28
Tang, Siew Tzuh, and Jane Dixon. "Instrument Translation and Evaluation of Equivalence and Psychometric Properties: The Chinese Sense of Coherence Scale." Journal of Nursing Measurement 10, no. 1 (May 2002): 59–76. http://dx.doi.org/10.1891/jnum.10.1.59.52544.
Full textAbstract:
Translating well-established English instruments into target languages other than English and testing cross-cultural validity to prove that the same attributes are being measured in each cultural group or country are required before a multicultural or international study can be conducted and cross-cultural comparisons of study results can be applied. However, rigorous and systematic cross-cultural efforts to test the effectiveness of specific translation methods are rare. This article presents a model of translation processes and empirical validation of the translated instrument through description of the translation of a selected instrument—the Sense of Coherence (SOC) Scale—from English into Chinese. A cyclic process of forward translations, back translations, and expert evaluation of equivalence by bilingual and English speaking experts was conducted to achieve conceptual equivalence between the original and translated instruments. Empirical validation of the Chinese SOC scale data from a group of bilingual Chinese people demonstrated non-identical, but comparable, item and scale means and variances, internal consistency, and relationships to an external criterion. Advantages and difficulties of using multi-rounds, multi-experts, and multi-methods to establish translation equivalence and to validate the translated Chinese SOC scale are discussed.
29
Asimakoulas, Dimitris. "Towards a Model of Describing Humour Translation." Meta 49, no. 4 (January 13, 2005): 822–42. http://dx.doi.org/10.7202/009784ar.
Full textAbstract:
Abstract Being rooted in a specific cultural and linguistic context, humour can pose significant problems to translation. This paper will discuss data collected from films in the light of a suggested framework based on script theory of humour initially proposed by Attardo and specifically adapted here for subtitling. The data include such categories as wordplay, where a more ‘semiotic’ approach is employed, comparisons, parody, disparagement and register humour. These data were culled from two films translated into Greek: Airplane! (1980), directed by David Zucker and Jim Abrahams and The Naked Gun: From the Files of the Police Squad (1988), directed by David Zucker, which exhibit a great concentration of verbal humorous sequences and inventive puns. It will be suggested that there was leeway to creatively solve linguistically/culturally based translation problems, although inconsistencies were to be observed.
30
Zunoomy, M. S., and F. H. A. Shibly. "IMPACT OF MACHINE TRANSLATION ON UNIVERSITY STUDENTS’ TRANSLATION ABILITY: SOUTH EASTERN UNIVERSITY OF SRI LANKA AS A MODEL." Indonesian Journal of Social Research (IJSR) 3, no. 2 (August 30, 2021): 121–27. http://dx.doi.org/10.30997/ijsr.v3i2.138.
Full textAbstract:
Machine Translation (MT) is a unique tool in the field of translation. It is used all over the world. In accordance with, the undergraduates who are following Specialization in Linguistics and Translation at the department of Arabic language from South Eastern University of Sri Lanka face obstacles when they translate between Tamil, English & Arabic. Due to, they focus on MT to translate. According to this, the current research aims to identify the impact of MT on their translation activities. This research uses the descriptive analysis methodology. Primary data were collected from observation and questionnaire that was given to the undergraduates who are following Specialization in Linguistics and Translation at the department of Arabic language from South Eastern University of Sri Lanka in the academic year 2018/2019. Secondary data were gathered from research papers, books, research articles, and websites. The findings of this research declares that the undergraduates in the selected area have desires in the translation field and face the obstacles when translating. Thus, they try to use MT for getting accurate translation, idea of the source text, understanding the unknown words. Due to it, they couldn’t omit the MT in their translation activities. Because, it is easy to use and helps to save the time. At the time, they assume that non-use of MT will effect negatively in their translation ability.
31
Kotambunan, Florence E. "Strategi Penambahan dan Penghilangan Makna: Kasus Penerjemahan Teks Hukum Bisnis dari Bahasa Inggris ke bahasa Indonesia." Paradigma, Jurnal Kajian Budaya 7, no. 2 (January 24, 2018): 199. http://dx.doi.org/10.17510/paradigma.v7i2.178.
Full textAbstract:
<span>To show how important the equivalence of ST and TT in business law text is the main goal of this research. Besides that, the benefit of this research is to motivate a translator to become more critical and accurate in producing more quality translations. The results of research can be utilized as a benchmark to conduct further research in a similar study. The library research and field method are commonly administered in translation research. And the technique of analyzing data exerts comparative and causal model between the Source Text (ST) and the Target Text (TT). It is determined that the approach is pursuant to ST and TT both Nord Extra-textual and Intra-textual factors. In addition, the research scope is limited to addition and omission, the strategy of translating in word, phrase or term, clause, and sentence. The conformity of ST and TT to Grice’s maxim is also analyzed as the effect of employing strategy of translating, the addition and omission. These research findings, firstly verifies that extra-textual and intra-textual factors are mostly integrated in achieving equivalence of ST and TT. Secondly, it is also discovered that translation error of deviation, addition, and omission of meaning as the impact of translating strategy application, addition and omission. Thirdly, it is acquired that writing business law text in English is more conventional compared to Indonesian. In conclusion, the translation of business law text complies with faithful and conventional main criteria. Moreover, not only the mastery of other knowledge related but also good collaboration are required between concerned parties and determined the success of translation practice. Therefore, it is wished for these results and implications could be referred to proceed with comprehensive research in other legal texts. Hereafter, it is also advised that proofreading is performed more than twice to minimalize misinterpretation. </span>
32
Maya-Yescas, R., E. León-Becerril, and D. Salazar-Sotelo. "Translation of MAT Kinetic Data to Model Industrial Catalytic Cracking Units." Chemical Engineering & Technology 27, no. 7 (July 2004): 777–80. http://dx.doi.org/10.1002/ceat.200401971.
Full text
33
Atzeni, Paolo, Luigi Bellomarini, Francesca Bugiotti, Fabrizio Celli, and Giorgio Gianforme. "A runtime approach to model-generic translation of schema and data." Information Systems 37, no. 3 (May 2012): 269–87. http://dx.doi.org/10.1016/j.is.2011.11.003.
Full text
34
Rijnbeek, Peter R. "Converting to a Common Data Model: What is Lost in Translation?" Drug Safety 37, no. 11 (September 4, 2014): 893–96. http://dx.doi.org/10.1007/s40264-014-0221-4.
Full text
35
Ye, Yanping. "Translation Mechanism of Neural Machine Algorithm for Online English Resources." Complexity 2021 (April 5, 2021): 1–11. http://dx.doi.org/10.1155/2021/5564705.
Full textAbstract:
At the level of English resource vocabulary, due to the lack of vocabulary alignment structure, the translation of neural machine translation has the problem of unfaithfulness. This paper proposes a framework that integrates vocabulary alignment structure for neural machine translation at the vocabulary level. Under the proposed framework, the neural machine translation decoder receives external vocabulary alignment information during each step of the decoding process to further alleviate the problem of missing vocabulary alignment structure. Specifically, this article uses the word alignment structure of statistical machine translation as the external vocabulary alignment information and introduces it into the decoding step of neural machine translation. The model is mainly based on neural machine translation, and the statistical machine translation vocabulary alignment structure is integrated on the basis of neural networks and continuous expression of words. In the model decoding stage, the statistical machine translation system provides appropriate vocabulary alignment information based on the decoding information of the neural machine translation and recommends vocabulary based on the vocabulary alignment information to guide the neural machine translation decoder to more accurately estimate its vocabulary in the target language. From the aspects of data processing methods and machine translation technology, experiments are carried out to compare the data processing methods based on language model and sentence similarity and the effectiveness of machine translation models based on fusion principles. Comparative experiment results show that the data processing method based on language model and sentence similarity effectively guarantees data quality and indirectly improves the algorithm performance of machine translation model; the translation effect of neural machine translation model integrated with statistical machine translation vocabulary alignment structure is compared with other models.
36
Pham, Hai, Paul Pu Liang, Thomas Manzini, Louis-Philippe Morency, and Barnabás Póczos. "Found in Translation: Learning Robust Joint Representations by Cyclic Translations between Modalities." Proceedings of the AAAI Conference on Artificial Intelligence 33 (July 17, 2019): 6892–99. http://dx.doi.org/10.1609/aaai.v33i01.33016892.
Full textAbstract:
Multimodal sentiment analysis is a core research area that studies speaker sentiment expressed from the language, visual, and acoustic modalities. The central challenge in multimodal learning involves inferring joint representations that can process and relate information from these modalities. However, existing work learns joint representations by requiring all modalities as input and as a result, the learned representations may be sensitive to noisy or missing modalities at test time. With the recent success of sequence to sequence (Seq2Seq) models in machine translation, there is an opportunity to explore new ways of learning joint representations that may not require all input modalities at test time. In this paper, we propose a method to learn robust joint representations by translating between modalities. Our method is based on the key insight that translation from a source to a target modality provides a method of learning joint representations using only the source modality as input. We augment modality translations with a cycle consistency loss to ensure that our joint representations retain maximal information from all modalities. Once our translation model is trained with paired multimodal data, we only need data from the source modality at test time for final sentiment prediction. This ensures that our model remains robust from perturbations or missing information in the other modalities. We train our model with a coupled translationprediction objective and it achieves new state-of-the-art results on multimodal sentiment analysis datasets: CMU-MOSI, ICTMMMO, and YouTube. Additional experiments show that our model learns increasingly discriminative joint representations with more input modalities while maintaining robustness to missing or perturbed modalities.
37
Ko, Sang-Ki, Chang Jo Kim, Hyedong Jung, and Choongsang Cho. "Neural Sign Language Translation Based on Human Keypoint Estimation." Applied Sciences 9, no. 13 (July 1, 2019): 2683. http://dx.doi.org/10.3390/app9132683.
Full textAbstract:
We propose a sign language translation system based on human keypoint estimation. It is well-known that many problems in the field of computer vision require a massive dataset to train deep neural network models. The situation is even worse when it comes to the sign language translation problem as it is far more difficult to collect high-quality training data. In this paper, we introduce the KETI (Korea Electronics Technology Institute) sign language dataset, which consists of 14,672 videos of high resolution and quality. Considering the fact that each country has a different and unique sign language, the KETI sign language dataset can be the starting point for further research on the Korean sign language translation. Using the KETI sign language dataset, we develop a neural network model for translating sign videos into natural language sentences by utilizing the human keypoints extracted from the face, hands, and body parts. The obtained human keypoint vector is normalized by the mean and standard deviation of the keypoints and used as input to our translation model based on the sequence-to-sequence architecture. As a result, we show that our approach is robust even when the size of the training data is not sufficient. Our translation model achieved 93.28% (55.28%, respectively) translation accuracy on the validation set (test set, respectively) for 105 sentences that can be used in emergency situations. We compared several types of our neural sign translation models based on different attention mechanisms in terms of classical metrics for measuring the translation performance.
38
Alshammari, Jaber Nashi M. "Analyzing Arabic Translation Methods of English Similes: A Case Study of The Old Man and The Sea by Ernest Hemingway." Theory and Practice in Language Studies 6, no. 3 (March 21, 2016): 485. http://dx.doi.org/10.17507/tpls.0603.05.
Full textAbstract:
Simile is one of the most important literary devices. It is widely used as a figure of speech in literary works. However, simile can pose significant challenges in literary translation since different languages might use and interpret similes differently. The present research aims at investigating the translation strategies employed in Arabic to render English similes in a literary text. The translation model proposed by Pierini (2007) is utilized as a framework of this study. The researcher selected "The Old Man and The Sea" novel by Ernest Hemingway and its two Arabic translations as a case study. The novel's two Arabic translations are by The United Publishers referred to later as target text 1 (TT1) and Zyad Zakaria referred to later as target text 2 (TT2). First, the researcher randomly collected 40 similes as the study data. Then, their Arabic translations are identified. Next, the data is compared and analyzed to determine their translation techniques. After analysis, the research found that literal translation is a prominent strategy in rendering English similes to Arabic.
39
Siregar, Masitowarni, T. Silvana Sinar, Amrin Saragih, and Syahron Lubis. "Need Analysis for Developing Translation’s Textbook base on TEFL Pedagogical Purpose in Indonesia: English Teachers’ Perspectives." Advances in Language and Literary Studies 9, no. 3 (June 30, 2018): 81. http://dx.doi.org/10.7575/aiac.alls.v.9n.3p.81.
Full textAbstract:
The purpose of this study was to find out the Prospective English teachers’ perspective on the development of translation’s textbook base on Teaching English as a Foreign Language (TEFL) Pedagogical purpose. Therefore, the quantitative research was applied to gather the data. By using the questionnaire with Likert Scale, the quantitative data were collected from 100 prospective English teachers in Indonesia. As a result, from nine aspects questioned to the prospective English teachers, most of them (92%) said that the translation’s textbook base on TEFL Pedagogical purpose was significantly needed. Therefore, the English Education Study Program (EESP) in Indonesia should use translation for teaching English purpose not for professional career in translation. Therefore, Translation as a TEFL purpose model is the most suitable model to be applied in EESP in Indonesia.
40
Ren, Beibei. "The use of machine translation algorithm based on residual and LSTM neural network in translation teaching." PLOS ONE 15, no. 11 (November 19, 2020): e0240663. http://dx.doi.org/10.1371/journal.pone.0240663.
Full textAbstract:
With the rapid development of big data and deep learning, breakthroughs have been made in phonetic and textual research, the two fundamental attributes of language. Language is an essential medium of information exchange in teaching activity. The aim is to promote the transformation of the training mode and content of translation major and the application of the translation service industry in various fields. Based on previous research, the SCN-LSTM (Skip Convolutional Network and Long Short Term Memory) translation model of deep learning neural network is constructed by learning and training the real dataset and the public PTB (Penn Treebank Dataset). The feasibility of the model’s performance, translation quality, and adaptability in practical teaching is analyzed to provide a theoretical basis for the research and application of the SCN-LSTM translation model in English teaching. The results show that the capability of the neural network for translation teaching is nearly one times higher than that of the traditional N-tuple translation model, and the fusion model performs much better than the single model, translation quality, and teaching effect. To be specific, the accuracy of the SCN-LSTM translation model based on deep learning neural network is 95.21%, the degree of translation confusion is reduced by 39.21% compared with that of the LSTM (Long Short Term Memory) model, and the adaptability is 0.4 times that of the N-tuple model. With the highest level of satisfaction in practical teaching evaluation, the SCN-LSTM translation model has achieved a favorable effect on the translation teaching of the English major. In summary, the performance and quality of the translation model are improved significantly by learning the language characteristics in translations by teachers and students, providing ideas for applying machine translation in professional translation teaching.
41
Zhang, Lin. "Web-Based Data Translation for a Resource Management Grid." Advanced Materials Research 989-994 (July 2014): 4379–82. http://dx.doi.org/10.4028/www.scientific.net/amr.989-994.4379.
Full textAbstract:
Grid environment as the cross field of semantic web and grid computing, has received significant attention recently. Our workgroup develops a resource management semantic grid named RM-Grid to enhance information process accuracy and efficiency in heterogeneity environment. This paper describes the architecture of RM-grid, mainly introduces the data translation service of RM-Grid. It puts forward an ontology based common semantic model and an XML based data model, presents the data translation algorithm between different data sources and an adapter framework for accessing distributed heterogeneous resources as well.
42
Dimelow, Richard J., and Stephen J. Wilkinson. "Control of translation initiation: a model-based analysis from limited experimental data." Journal of The Royal Society Interface 6, no. 30 (June 20, 2008): 51–61. http://dx.doi.org/10.1098/rsif.2008.0221.
Full textAbstract:
We have built a detailed kinetic model of translation initiation in yeast and have used a novel approach to determine the flux controlling steps based on limited experimental data. An efficient parameter estimation method was adapted in order to fit the most uncertain parameters (rate constants) to in vivo measurements in yeast. However, it was found that there were many other sets of plausible parameter values that also gave a good fit of the model to the data. We therefore used random sampling of this uncertain parameter space to generate a large number of diverse fitted parameter sets. A compact characterization of these parameter sets was provided by considering flux control. In particular, we suggest that the rate of translation initiation is most strongly influenced by one of two reactions: either the guanine nucleotide exchange reaction involving initiation factors eIF2 and eIF2B or the assembly of the multifactor complex from its constituent protein/tRNA containing complexes. It is hoped that the approach presented in this paper will add to our understanding of translation initiation pathway and can be used to identify key system-level properties of other biochemical processes.
43
Lee, Yong Joo, Seong-Jong Joo, and Hong Gyun Park. "An application of data envelopment analysis for Korean banks with negative data." Benchmarking: An International Journal 24, no. 4 (May 2, 2017): 1052–64. http://dx.doi.org/10.1108/bij-02-2016-0023.
Full textAbstract:
Purpose The purpose of this paper is to measure the comparative efficiency of 18 Korean commercial banks under the presence of negative observations and examine performance differences among them by grouping them according to their market conditions. Design/methodology/approach The authors employ two data envelopment analysis (DEA) models such as a Banker, Charnes, and Cooper (BCC) model and a modified slacks-based measure of efficiency (MSBM) model, which can handle negative data. The BCC model is proven to be translation invariant for inputs or outputs depending on output or input orientation. Meanwhile, the MSBM model is unit invariant in addition to translation invariant. The authors compare results from both models and choose one for interpreting results. Findings Most Korean banks recovered from the worst performance in 2011 and showed similar performance in recent years. Among three groups such as national banks, regional banks, and special banks, the most special banks demonstrated superb performance across models and years. Especially, the performance difference between the special banks and the regional banks was statistically significant. The authors concluded that the high performance of the special banks was due to their nationwide market access and ownership type. Practical implications This study demonstrates how to analyze and measure the efficiency of entities when variables contain negative observations using a data set for Korean banks. The authors have tried two major DEA models that are able to handle negative data and proposed a practical direction for future studies. Originality/value Although there are research papers for measuring the performance of banks in Korea, all of the papers in the topic have studied efficiency or productivity using positive data sets. However, variables such as net incomes and growth rates frequently include negative observations in bank data sets. This is the first paper to investigate the efficiency of bank operations in the presence of negative data in Korea.
44
Et. al., Muhammad Hasyim,. "Human-Robots And Google Translate: A Case Study Of Translation Accuracy In Translating French-Indonesian Culinary Texts." Turkish Journal of Computer and Mathematics Education (TURCOMAT) 12, no. 4 (April 11, 2021): 1194–202. http://dx.doi.org/10.17762/turcomat.v12i4.633.
Full textAbstract:
Google Translate (GT) is the most widely used translator application in the world. The function of GT is not merely as tools but has become a means in personal communication, learning and business matters. This paper aims to examine the GT accuracy in translating culinary texts. This paper used a semiotic approach to analyze the equivalence of GT from the source language to the target language. The data source as the object of study is French culinary texts retrieved from the internet. It can be concluded that the accuracy of GT in culinary texts lies in words, phrases and sentences. GT does not yet have good accuracy in translating culture between source and target languages. In the semiotic perspective, the GT translation model is the translation from the source language form to the target language form (signifier) by maintaining the concept (meaning). GT does not yet have a translation feature in a cultural context. The advancement in information technology shows that GT, as a human-robot, can replace human translators.
45
Arhire, Mona. "Cătălina Iliescu-Gheorghiu: a polysystemic model for the comparative analysis of drama from the perspective of descriptive translation studies." Swedish Journal of Romanian Studies 3, no. 1 (April 17, 2020): 259–64. http://dx.doi.org/10.35824/sjrs.v3i1.20438.
Full textAbstract:
This review presents a recently published book authored by Cătălina Iliescu Gheorghiu, an academic actively involved in Romanian studies and a translator of Romanian literature. As the title suggests, it is a study that falls under the scope of Descriptive Translation Studies implying the polysystemic model posited by Lambert and Van Gorp for the comparative analysis of drama. The corpus under scrutiny is made up utterances extracted from the play A treia țeapă (The Third Stake) by Marin Sorescu and the corresponding utterances from two of its translations into English. The analytical part is backed up by a solid theoretical framework with its latter section lending the overall structure of the analysis. The categories subject to investigation are (i) preliminary data, (ii) the macro-level structures, (iii) the micro-level structures and (iv) the systemic context. The methodology experimented with drama translation and the findings deriving from it have proved their validity and are valuable input for other similar and possibly more comprising research that can use these findings as hypotheses to be tested further.
46
Vulić, Ivan, and Marie-Francine Moens. "Bilingual Distributed Word Representations from Document-Aligned Comparable Data." Journal of Artificial Intelligence Research 55 (April 12, 2016): 953–94. http://dx.doi.org/10.1613/jair.4986.
Full textAbstract:
We propose a new model for learning bilingual word representations from non-parallel document-aligned data. Following the recent advances in word representation learning, our model learns dense real-valued word vectors, that is, bilingual word embeddings (BWEs). Unlike prior work on inducing BWEs which heavily relied on parallel sentence-aligned corpora and/or readily available translation resources such as dictionaries, the article reveals that BWEs may be learned solely on the basis of document-aligned comparable data without any additional lexical resources nor syntactic information. We present a comparison of our approach with previous state-of-the-art models for learning bilingual word representations from comparable data that rely on the framework of multilingual probabilistic topic modeling (MuPTM), as well as with distributional local context-counting models. We demonstrate the utility of the induced BWEs in two semantic tasks: (1) bilingual lexicon extraction, (2) suggesting word translations in context for polysemous words. Our simple yet effective BWE-based models significantly outperform the MuPTM-based and context-counting representation models from comparable data as well as prior BWE-based models, and acquire the best reported results on both tasks for all three tested language pairs.
47
Song, Hye-Jeong, Hong-Ki Kim, Jong-Dae Kim, Chan-Young Park, and Yu-Seop Kim. "Inter-Sentence Segmentation of YouTube Subtitles Using Long-Short Term Memory (LSTM)." Applied Sciences 9, no. 7 (April 11, 2019): 1504. http://dx.doi.org/10.3390/app9071504.
Full textAbstract:
Recently, with the development of Speech to Text, which converts voice to text, and machine translation, technologies for simultaneously translating the captions of video into other languages have been developed. Using this, YouTube, a video-sharing site, provides captions in many languages. Currently, the automatic caption system extracts voice data when uploading a video and provides a subtitle file converted into text. This method creates subtitles suitable for the running time. However, when extracting subtitles from video using Speech to Text, it is impossible to accurately translate the sentence because all sentences are generated without periods. Since the generated subtitles are separated by time units rather than sentence units, and are translated, it is very difficult to understand the translation result as a whole. In this paper, we propose a method to divide text into sentences and generate period marks to improve the accuracy of automatic translation of English subtitles. For this study, we use the 27,826 sentence subtitles provided by Stanford University’s courses as data. Since this lecture video provides complete sentence caption data, it can be used as training data by transforming the subtitles into general YouTube-like caption data. We build a model with the training data using the LSTM-RNN (Long-Short Term Memory – Recurrent Neural Networks) and predict the position of the period mark, resulting in prediction accuracy of 70.84%. Our research will provide people with more accurate translations of subtitles. In addition, we expect that language barriers in online education will be more easily broken by achieving more accurate translations of numerous video lectures in English.
48
Pfordresher, Peter Q., Andrea R. Halpern, and Emma B. Greenspon. "A Mechanism for Sensorimotor Translation in Singing." Music Perception 32, no. 3 (February 1, 2015): 242–53. http://dx.doi.org/10.1525/mp.2015.32.3.242.
Full textAbstract:
We propose a new framework to understand singing accuracy, based on multi-modal imagery associations: the MMIA model. This model is based on recent data suggesting a link between auditory imagery and singing accuracy, evidence for a link between imagery and the functioning of internal models for sensorimotor associations, and the use of imagery in singing pedagogy. By this account, imagery involves automatic associations between different modalities, which in the present context comprise associations between pitch height and the regulation of vocal fold tension. Importantly, these associations are based on probabilistic relationships that may vary with respect to their precision and accuracy. We further describe how this framework may be extended to multi-modal associations at the sequential level, and how these associations develop. The model we propose here constitutes one part of a larger architecture responsible for singing, but at the same time is cast at a general level that can extend to multi-modal associations outside the domain of singing.
49
Zasiekin, Serhii. "Understanding translation universals." Babel. Revue internationale de la traduction / International Journal of Translation 62, no. 1 (May 19, 2016): 122–34. http://dx.doi.org/10.1075/babel.62.1.07zas.
Full textAbstract:
Translation S- and T-universals have been widely discussed in Translation Studies and their psycholinguistic study has been among the priority topics today. The article is focused on the study of translation ‘S- universals’ (Chesterman 2004) and is based both on the psycholinguistic model of literary translation, which combines today’s neuroscience theories of cognitivism and connectionism, and on the experimental data that demonstrate its validity. The model is resulted in a series of experiments held with undergraduate students of translation department. The results of the study proved the idea of existing specific procedural and discursive S-universals in literary translation. As the empirical data showed, these universals maintain the status of common strategies depending on translator’s cognitive style (analytical or synthetic) and his dominant channel (visual, auditory, kinesthetic) of source text perception.
50
Matvieieva, Svitlana. "TERM AS A KEY UNIT OF LEGAL PARALLEL CORPUS: TRANSLATION ASPECT." Studia Linguistica, no. 15 (2019): 155–64. http://dx.doi.org/10.17721/studling2019.15.155-164.
Full textAbstract:
The article deals with the development of a corpus approach to the analysis and translation of law terminology. The object of the study is terminology of legal discourse as a semantic phenomenon, the subject is the specifics of the translation of a legal term using the corpus approach. The source of the research material is a parallel English-Ukrainian corpus of legal texts compiled by the author of the article from the decisions of the European Court in English and their translations into Ukrainian with a volume of about 275,000 words (English component). To test the productivity of applying the corpus approach to the translation of a term, the article analyzes the legal term “a perpetrator”, which is a part of the cognitive model LEGAL PROCESS – AGENT / SUBJECT OF ACTION, and its translation into Ukrainian. The analysis of the term in the dictionaries of various types and the comparison of these data with the materials of the corpus of parallel professional (legal) texts carried out in the article demonstrate the absence of certain translation variants for the term being analyzed in the dictionaries. It is important to study the possibilities of corpora of parallel texts of professional metalanguages as sources for clarifying and choosing the most accurate variants for translating terms, enriching terminological and translation dictionaries and developing thesauri of various subject areas.