Machine Translation Research Articles

ANALYSIS OF GOOGLE TRANSLATE’S TRANSLATION AND THE AUTHOR’S FROM ALBANIAN TO SERBIAN LANGUAGE

Literary translation, as one of the most complex forms of translation, requires a special approach and attention, given that it involves not only conveying meanings but also preserving the style of the original text. Although the translation of texts using various online tools is becoming increasingly common today, the question arises - are these tools advanced enough, and can they replace the translation done by human authors? As our corpus material, we used specific excerpts from the first chapter of Ismail Kadare’s novel ‘The Palace of Dreams,’ which we translated using Google Translate and compared with the Serbian translation of this novel by Shkëlzen Maliqi. The aim of the paper is to examine the reliability of machine translation, i.e. analysing the mistakes it makes, compared to the translation of competent and professional translators. The comparison process is based on different linguistic disciplines, and we have come to the conclusions that Google Translate has not yet improved enough to become the main tool in the translation of literary texts, where professional translators still play a key role. Of course, the question arises - whether it would perform better in translating other types of literary texts or in translating different language groups, considering its diverse corpus of literary texts in its memory?

Developing and Deploying End‐to‐End Machine Learning Systems for Social Impact: A Rubric and Practical Artificial Intelligence Case Studies From African Contexts

ABSTRACTArtificial intelligence (AI) and machine learning have demonstrated the potential to provide solutions to societal challenges, for example, automated crop diagnostics for smallholder farmers, environmental pollution modelling and prediction for cities and machine translation systems for languages that enable information access and communication for segments of the population who are unable to speak or write official languages, among others. Despite the potential of AI, the practical and technical issues related to its development and deployment in the African context are the least documented and understood. The development and deployment of AI for social impact systems in the developing world present new intricacies and requirements emanating from the unique technology and social ecosystems in these settings. This paper provides a rubric for developing and deploying AI systems for social impact with a focus on the African context. The rubric is derived from the analysis of a series of selected real‐world case studies of AI applications in Africa. We assessed the selected AI case studies against the proposed rubric. The rubric and examples of AI applications presented in this paper are expected to contribute to the development and application of AI systems in other African contexts.

A deaf-centred art-science approach to community engagement with sign language technologies

Sign Language Machine Translation (SLMT) is an emerging technology that will primarily impact deaf and hard of hearing communities. Historically, these technologies have been developed by hearing scientists, without input from deaf researchers or community representatives. SignON was a three-year project exploring the application of SLMT, and using a co-creation approach which was led by the European Union of the Deaf. Here we describe an art-science method to engage deaf, hard of hearing, and hearing audiences with SLMT through theatre and performance. We also reflect on the insights, perspectives, and feedback shared by audience members.

The Relationship Between the Role of Madrasah Heads, Work Motivation, and Professional Competence with the Performance of State Madrasah Tsanawiyah Teachers

This study aims to analyse: (1) the relationship between the role of the principal of religious school in Malang County and teachers' performance in MD as agreed by teachers respectively; machine translation teachers in Malang County relationship; (3) there is a relationship between the ability and performance of national machine translation teachers in Malang City; and the performance of machine translation teachers in the social network of Malang County. This study used various descriptive and correlational analysis methods to determine the relationship between the variables. The results of this research show that: (1) The role of religious school has a positive relationship with the achievement of teachers and sig scores. If equal to 0.000 < 0.05, the calculated r value is 0.323 > r table is 0.146; If equal to 0.000 < 0.05, the calculated r value is 0.713 > r table is 0.146; If equal to 0.000 < 0.05, the calculated r value is 0.812> r table is 0.146; (4) The role of the religious school principal, support work and ability are related to the success of teachers and SIG scores. 0.000 < 0.05, the strength of the relationship is 0.829, which means the relationship is in the excellent category

Dap-SiMT: divergence-based adaptive policy for simultaneous machine translation

Dap-SiMT: divergence-based adaptive policy for simultaneous machine translation

Image-to-image machine translation enables computational defogging in real-world images

Image-to-image machine translation enables computational defogging in real-world images

Machine Translation Shifts on The Meaning Equivalence of Culture Sentence and Illocutionary Speech Acts: Back-Translation

Machine Translation Shifts on The Meaning Equivalence of Culture Sentence and Illocutionary Speech Acts: Back-Translation

Gender Issues between Gemini and ChatGPT: The Case of English-Arabic Translation

The study focuses on the gender-related issues that face English-Arabic machine translation. It aims to investigate and evaluate gender accuracy in the translations provided by two prominent large language models, Gemini and ChatGPT, recognizing the rich morphological system of Arabic that includes gender marking. The researchers develops a test suite to evaluate gender accuracy in the translation outputs of Gemini and ChatGPT. The evaluation is performed by two professional annotators. That is followed by an analysis of the patterns of the gender-related issues that appear in the translation outputs of the models under study. The results show that Gemini outperformed ChatGPT in almost every aspect when it comes to gender-related translation issues. Both the number of the annotated issues as well as the gender accuracy evaluation came in favor of Gemini. The study introduced different patterns of gender-related translation issues. It also provides recommendations for future research.

Towards explainable fake news detection and automated content credibility assessment: Polish internet and digital media use-case

Detecting fake news and disinformation is a challenging and complex task. With the proliferation of the transformer architectures, researchers and practitioners have started actively using them to combat this phenomenon. In that regard, different approaches have been proposed by the community. Unfortunately, a significant number of solutions have started to treat misinformation in a similar way as other natural language processing challenges (e.g., sentiment classification). However, the reality is more complex, and the fake news detection problem often require additional context that may be essential to provide a complete assessment. Another pitfall in combating fake news is treating it as a binary classification challenge (e.g., fake vs. legitimate news). Combining such a simplified response with a deep black-box neural system effectively impacts the interpretability and eventual trustworthiness of the detection system. In this paper, the phenomenon of fake news is approached from a multi-factor perspective. Therefore, the proposed innovative approach does not limit news content rating to a one-dimensional binary response. In particular, the authors utilize a vast spectrum of processing techniques, including text embedding, machine translation, and various users’ preferences to provide content assessment. The results are presented using a Polish digital media use-case.

On the Application of Machine Translation in Legal Terminology Translation

On the Application of Machine Translation in Legal Terminology Translation

Axiological and cultural codes of verbal semantics in the context of machine translation (based on the verbs of the Russian, Tatar and English languages)

Axiological and cultural codes of verbal semantics in the context of machine translation (based on the verbs of the Russian, Tatar and English languages)

Automatic rating method based on deep transfer learning for machine translation considering contextual semantic awareness

With the acceleration of globalization, machine translation (MT) plays an increasingly prominent role in cross-language communication. However, how to evaluate the quality of machine translation, especially considering the nuances in different semantic contexts, remains a challenge. This paper proposes an automatic scoring model for machine translation quality based on deep transfer learning, which aims to accurately perceive and evaluate the quality of translated texts in different semantic contexts. Firstly, a pre-trained deep neural network model is used to extract the semantic feature representation of the sentences in the source language and the target language, so as to capture the semantic information of the sentences. Then, deep transfer learning is used to map the semantic features of source language and target language into the shared feature space. By sharing feature space, an effective relationship is established between the semantic representations of two languages, so as to achieve cross-language quality evaluation. The experimental results show that the model has made significant progress in evaluating the quality of machine translation. Compared with traditional methods, this model can evaluate translation quality more accurately and has better generalization ability. The automatic scoring model of machine translation quality based on deep transfer learning not only improves the accuracy and efficiency of machine translation quality assessment, but also provides new ideas and methods for future research.

COVERT SIMULTANEOUS POST-EDITING IN ONLINE ASSESSMENT OF STUDENTS’ SIGHT TRANSLATION

This first attempt aims to determine the extent of students’ covert use of machine translation (MT) in the online assessment of their sight translation, the strategies of such use, and its signs. The study is based on the analysis of target texts (TT) of specialised online sight translation from Ukrainian into English by 13 BA and 10 MA students. The procedure involved the comparison of the students' TTs with their MT counterparts. Signs of covert MT were found in 46% of the BA and 30% of the MA students’ translations. The main method of this covert MT use is "simultaneous post-editing", i.e., the immediate oral post-editing of the MT text generated by the students on their screens and hidden from the assessor, while they deliver their supposedly original TTs. Simultaneous post-editing strategies range from replacing individual lexemes with their synonyms, adding and deleting elements, changing the syntactic functions of words or phrases, rearranging sentence fragments, transforming their structure, to applying several of these strategies simultaneously. Other methods of concealment include alternating MT systems in translating the same source text, as well as artificially slowing down the process of reading the TT from the screen, accompanied by pauses in the relevant text fragments to perform certain mental operations. In order to increase objectivity, the author recommends a delayed assessment of students' online interpreting recordings. The research perspective is to study the didactic potential of simultaneous post-editing as a procedure for developing general interpreting skills.

LLMs-based machine translation for E-commerce

Large language models(LLMs) have shown promising performance for various downstream tasks, especially machine translation. However, LLMs and Specialized Translation Models (STMs) are designed to handle general translation needs, they are not well-suited for domains with specialized terms and writing styles, such as e-commerce, legal, and medicine. In the e-commerce domain, the text often contains many domain-specific terms and keyword-stacked structures, leading to poor translation quality with existing NMT methods. To tackle these problems, we have collected two resources specifically for the e-commerce domain, including aligned Chinese-English bilingual terms and parallel corpus from real e-commerce scenarios for model fine-tuning. We propose an LLMs-based E-commerce machine translation approach(LEMT) which includes LLMs utilization, e-commerce resources collection, and tokenizer optimization. We conduct two-stage fine-tuning and self-contrastive enhancement based on general LLMs to enable the model to learn translation features in the e-commerce domain. Through comprehensive evaluations on real e-commerce titles, our LEMT methodology demonstrates superior translation quality and robustness, outperforming leading NMT models such as NLLB, LLaMA, and even GPT-4.

Domain Adaptation for Arabic Machine Translation: Financial Texts as a Case Study

Domain Adaptation for Arabic Machine Translation: Financial Texts as a Case Study

Comparative Analysis of Online Translators in the Machine Translation System

The article presents a comparative analysis of various online translators in the machine translation system. The author of the study focused on finding out the effectiveness and accuracy of the translation, identifying the strengths and weaknesses of the three most common online translators: Google Translate, DeepL and Microsoft Translator. The article begins with an overview of current trends in the field of machine translation and the need for online translations based on this context. The research methodology is selected in the article, in particular, the selection of individual online translators for comparison and the criteria for evaluating their work. The researcher analyzes the quality of translations for various types of texts, including common phrases, technical terms, complex sentences, etc. Attention is also paid to different language pairs and ensuring the availability of rarer languages in electronic translation. The article reveals such aspects as accuracy of translation, speed of work, recognition of context and idiomatic expressions, availability of additional functions and capabilities that ensure better translation quality. The author thoroughly highlights the strengths and weaknesses of each system, provides clear information about the use of integrations, available functions and additional capabilities of each translator. In the final part of the article, the author provides conclusions and recommendations regarding the use of online translators in the machine translation system. The research highlights that each electronic translator has its own advantages and limitations, and the choice depends on the specific needs of the user. In addition, the author emphasizes the need for constant updating and improvement of online translators to increase their accuracy and efficiency. This comparative analysis will help simplify the selection of the optimal translation software for users who require high-quality machine translation for their professional activities or everyday needs.

КӨРКӨМ КОТОРМОДО САНАРИПТИК ТЕХНОЛОГИЯЛАРДЫ КОЛДОНУУ

In today's world, digital technologies play a key role in the field of translation, especially in a formal business context. However, their use in the translation of fiction remains less common. This article explores the possibilities of using digital methods in pre-translation activities and evaluates their effectiveness in the context of translating literary works. Special attention is paid to the issues of introducing digital technologies into the work of a translator of literary texts and their potential application in teaching. Using the example of the translation of Kyrgyz poetry, the problems arising from the use of digital means, including machine translation, are considered. This article offers an analysis of the possibilities and limitations of digital technologies in the context of translating literary texts in order to improve the quality and speed of translation. Digital technologies are actively being introduced into the work of a translator, especially in such areas as written translation in official business discourse. Digital technologies are used less often in the translation of fiction, but such techniques allow for pre-translation activities. In other words, it provides a literal translation of the text and helps to find the literal meaning of each word in the dictionary. However, some digital methods can be successfully used in the translation of literary works.

Building efficient and effective OpenQA systems for low-resource languages

Question answering (QA) is the task of answering questions posed in natural language with free-form natural language answers extracted from a given passage. In the OpenQA variant, only a question text is given, and the system must retrieve relevant passages from an unstructured knowledge source and use them to provide answers, which is the case in the mainstream QA systems on the Web. QA systems currently are mostly limited to the English language due to the lack of large-scale labeled QA datasets in non-English languages. In this paper, we show that effective, low-cost OpenQA systems can be developed for low-resource contexts. The key ingredients are (1) weak supervision using machine-translated labeled datasets and (2) a relevant unstructured knowledge source in the target language context. Furthermore, we show that only a few hundred gold assessment examples are needed to reliably evaluate these systems. We apply our method to Turkish as a challenging case study, since English and Turkish are typologically very distinct and Turkish has limited resources for QA. We present SQuAD-TR, a machine translation of SQuAD2.0, and we build our OpenQA system by adapting ColBERT-QA and retraining it over Turkish resources and SQuAD-TR using two versions of Wikipedia dumps spanning two years. We obtain a performance improvement of 24–32% in the Exact Match (EM) score and 22–29% in the F1 score compared to the BM25-based and DPR-based baseline QA reader models. Our results show that SQuAD-TR makes OpenQA feasible for Turkish, which we hope encourages researchers to build OpenQA systems in other low-resource languages. We make all the code, models, and the dataset publicly available at https://github.com/boun-tabi/SQuAD-TR.

A Novel Alignment-based Approach for PARSEVAL Measures

Abstract We propose a novel method for calculating PARSEVAL measures to evaluate constituent parsing results. Previous constituent parsing evaluation techniques were constrained by the requirement for consistent sentence boundaries and tokenization results, proving to be stringent and inconvenient. Our new approach handles constituent parsing results obtained from raw text, even when sentence boundaries and tokenization differ from the preprocessed gold sentence. Implementing this measure is our evaluation by alignment approach. The algorithm enables the alignment of tokens and sentences in the gold and system parse trees. Our proposed algorithm draws on the analogy of sentence and word alignment commonly used in machine translation (MT). To demonstrate the intricacy of calculations and clarify any integration of configurations, we explain the implementations in detailed pseudo-code and provide empirical proof for how sentence and word alignment can improve evaluation reliability.

A Global Features and Local Features-Jointly Fused Deep Semantic Learning Framework for Error Detection of Machine Translation

A Global Features and Local Features-Jointly Fused Deep Semantic Learning Framework for Error Detection of Machine Translation