Search

AbstractThis paper investigates the extent to which Arabic punctuation is rule-governed, with the aim of improving text comprehension, disambiguation, and machine translation. The study highlights the lack of systematic punctuation in Arabic written discourse, which may be attributed to difficulties in sentence boundary identification or inadequate differentiation between various conjunctions. The punctuation behavior of Arabic speakers is examined in relation to sentence boundary identification and the level of agreement among Arabic specialists is assessed. A quantitative analysis of paragraph and sentence lengths across genres, categories of writers, and in comparison to English is conducted using five corpora specifically compiled for this study. Additionally, a punctuation survey is carried out to evaluate specialists’ agreement on sentence boundary identification. The results indicate that writers of Arabic interpret punctuation rules differently and that Arabic punctuation practice is irregular. The study suggests that standardization of Arabic punctuation rules is necessary to facilitate comprehension and automatic text processing.

Background: Machine translation (MT) apps are used informally by healthcare professionals in many settings, especially where interpreters are not readily available. As MT becomes more accurate and accessible, it may be tempting to use MT more widely. Institutions and healthcare professionals need guidance on when and how these applications might be used safely and how to manage potential risks to communication. Objectives: Explore factors that may hinder or facilitate communication when using voice-to-voice MT. Design: Health professionals volunteered to use a voice-to-voice MT app in routine encounters with their patients. Both health professionals and patients provided brief feedback on the experience, and a subset of consultations were observed. Participants: Doctors, nurses, and allied health professionals working in the Primary Care Division of the Geneva University Hospitals, Switzerland. Main Measures: Achievement of consultation goals; understanding and satisfaction; willingness to use MT again; difficulties encountered; factors affecting communication when using MT. Key Results: Fourteen health professionals conducted 60 consultations in 18 languages, using one of two voice-to-voice MT apps. Fifteen consultations were observed. Professionals achieved their consultation goals in 82.7% of consultations but were satisfied with MT communication in only 53.8%. Reasons for dissatisfaction included lack of practice with the app and difficulty understanding patients. Eighty-six percent of patients thought MT-facilitated communication was easy, and most participants were willing to use MT in the future (73% professionals, 84% patients). Experiences were more positive with European languages. Several conditions and speech practices were identified that appear to affect communication when using MT. Conclusion: While professional interpreters remain the gold standard for overcoming language barriers, voice-to-voice MT may be acceptable in some clinical situations. Healthcare institutions and professionals must be attentive to potential sources of MT errors and ensure the conditions necessary for safe and effective communication. More research in natural settings is needed to inform guidelines and training on using MT in clinical communication. [ABSTRACT FROM AUTHOR]

The study aims to investigate the quality of neural machine translation when translating research paper abstracts from English to Arabic. It performs an error analysis and provides an evaluation of the quality of neural machine translation (NMT) represented by Google Translate and Microsoft Translator. The research team selects 25 English research paper abstracts in education from well-known Scopus scientific journals issued in English speaking countries. These abstracts are then translated into Arabic using both Google Translate and Microsoft Translator. The error analysis is based on the typology of errors introduced by Multidimensional Quality Metrics (MQM). A professional evaluation is also conducted using the Scalar Quality Metric evaluation (SQM) as proposed in Freitag (2021). The study finds that the translation outputs of academic texts like abstracts of education research papers are still not up to standards when translating English educational research abstracts into Arabic. There are various types of translation errors. However, there is a slight difference in translation quality and number of errors in favor of Google Translate compared to Microsoft Translator. However, it is included that NMT system still requires a lot of training, and more Arabic corpora need to be built. [ABSTRACT FROM AUTHOR]

Most methods for multi-domain adaptive neural machine translation (NMT) currently rely on mixing data from multiple domains in a single model to achieve multi-domain translation. However, this mixing can lead to imbalanced training data, causing the model to focus on training for the large-scale general domain while ignoring the scarce resources of specific domains, resulting in a decrease in translation performance. In this paper, we propose a multi-domain adaptive NMT method based on Domain Data Balancer (DDB) to address the problems of imbalanced data caused by simple fine-tuning. By adding DDB to the Transformer model, we adaptively learn the sampling distribution of each group of training data, replace the maximum likelihood estimation criterion with empirical risk minimization training, and introduce a reward-based iterative update of the bilevel optimizer based on reinforcement learning. Experimental results show that the proposed method improves the baseline model by an average of 1.55 and 0.14 BLEU (Bilingual Evaluation Understudy) scores respectively in English-German and Chinese-English multi-domain NMT. [ABSTRACT FROM AUTHOR]

This article analyzes texts that have been generated by machine translation (MT) and post-edited by native English-speaking trainee translators (English>Chinese) who are also Chinese language learners enrolled in a four-year undergraduate translation program. The project examines the work product of trainee translators to categorize 122 errors that are (un)noticed and (un)corrected by them. MT errors in the Accuracy category were best identified and corrected, followed by those in the Lexicon and Fluency categories. Trainee translators who were advanced language learners outperformed the intermediate group in MT error detection and correction, especially in the Lexicon category. This study sheds light upon the use of raw MT output as meaningful input for trainee translators who are in the process of learning Chinese. Its findings provide information regarding the type of exercises needed in language learning and translation training for students with different levels of language proficiency. [ABSTRACT FROM AUTHOR]

This article presents the results of a study focusing on the reception of a fictional story by Kurt Vonnegut translated from English into Catalan and Dutch in three conditions: machine translated, post-edited, and human translated. Participants (n = 223) rated the three conditions using three scales: narrative engagement, enjoyment, and translation reception. The results show that human translation had higher engagement, enjoyment, and translation reception in Catalan, compared to the post-edited and machine-translated translations. However, Dutch readers scored the post-edited translation higher than the human and machine translation, and the highest engagement and enjoyment scores were reported for the original English version. We hypothesize that when reading a fictional story in translation, not only are the condition and the quality of the translation key to understanding its reception, but also the participants' reading patterns, reading language, and, potentially, the status of the source language in their own societies. [ABSTRACT FROM AUTHOR]

This article presents a matrix of competence descriptors aimed at machine translation-oriented data literacy teaching. This competence matrix constitutes the didactics-facing side of the DataLitMT project, which develops learning resources for teaching relevant components of data literacy in their translation-specific form of professional machine translation (MT) literacy to BA and MA students in translation and specialised communication. After highlighting the increasing relevance of both professional MT literacy and data literacy in the context of Translation Studies and professional translation, the article presents and discusses a professional MT literacy framework and an MT-specific data literacy framework, which serve to structure the two frames of reference relevant to this article. Then, the article provides a detailed discussion of the competence matrix developed based on the two frameworks sketched previously. This discussion is intended to show how the individual dimensions and sub-dimensions of data literacy were linked to relevant (sub-)dimensions of professional MT literacy and translated into corresponding competence descriptors. To conclude, the article presents an example of a learning resource for MT-oriented data literacy teaching developed based on the descriptors of the competence matrix. [ABSTRACT FROM AUTHOR]

The recent paper highlights the impact of artificial intelligence on Machine Translation without the interaction of Humans. The use of Google Translator, Bing, Microsoft Translator, Systran Translate and Amazon Translate has become widely spread (CAT Tools). This study aims to reveal the contrast between Artificial Intelligence and Human Translation in the legal field. A hypothesis of the difference between Artificial Translation and Human Translation was raised. The concerns about the lack of a translator increased, and machine translation was selected as the most selected option. Local and foreign contracts were selected and subjected to Human and Machine translation. Strengths and weaknesses points were selected and analyzed. The previous studies in the legal translation field were considered. The results revealed the gap between human translation and machine translation, and human translation is dominant in the light of accuracy and the existence of legal language. The findings also focused on the Translators' experience and knowledge in the translation field. [ABSTRACT FROM AUTHOR]

The translation of culture-specific terms constitutes a major challenge for professional translators as it necessitates a thorough understanding of both the linguistic and cultural elements. With rapid technological advancement over the past few years, machine translation has enhanced translation quality. This study investigates the transability of cultural connotations in the Arabic-English translation of Naguib Mahfouz's Zuqāq al-Midaqq. A descriptive qualitative research design was adopted to achieve the intended goals of the study. The data comprised human and machine translations from Google Translate and ChatGPT. Through qualitative content analysis, the translations were compared for accuracy in transferring the cultural connotations prevalent in the Arabic source text. The findings revealed that the human translation showed greater naturalness and accuracy in rendering cultural connotations. Machine translation has struggled with rhetorical devices, idioms, and cultural nuances. The results also indicated that the AI-enhanced machine represented by ChatGPT captured the cultural elements more effectively than Google Translate. The study concluded that human expertise remains essential for the high-quality translation of literary works to maintain cultural significance. The findings can inform translator training and guide improvements to AIenhanced translation for literary texts. [ABSTRACT FROM AUTHOR]

The primary focus of artificial intelligence advancement is in machine translation; nonetheless, a prevalent issue persists in the form of imprecise translation. The current challenge faced by artificial intelligence is to effectively executing machine translation from extensive datasets. This research presents a BP neural method that aims to repeatedly analyse translation data and achieve optimisation in machine translation. The findings indicate that the use of BP neural network may enhance the dependability and precision of machine translation, with an accuracy rate over 84%. This performance surpasses that of the online translation approach. Hence, it can be inferred that the use of BP neural algorithms has the potential to fulfil the requirements of machine translation and enhance the precision of online translation conducted by humans. [ABSTRACT FROM AUTHOR]

With the development of science and technology and the rise of international trade, cross-border e-commerce platforms play an important role in international trade. While machine translation plays a crucial role in the introduction of goods in cross-border e-commerce, many machine translations have limitations, such as inaccurate descriptions and not conforming to Internet trends and cultural conflicts. Therefore, this study aims to investigate the characteristics of localization translation and apply them to the establishment of a machine translation corpus. By conducting questionnaires, this paper analyses the translation characteristics and user needs of cross-border e-commerce product introductions. The results found that product introduction translation needs to have the following characteristics: accurate description, comprehensive description, vivid description, concise description, and cross-cultural homogeneity. Based on the analysis, this study raises credible and valid suggestions and provides insights into building a machine translation corpus. [ABSTRACT FROM AUTHOR]

Background: Artificial intelligence (AI) is rapidly changing communication and technology-driven content creation and is also being used more frequently in education. Despite these advancements, AI-powered automated scoring in international large-scale assessments (ILSAs) remains largely unexplored due to the scoring challenges associated with processing large amounts of multilingual responses. However, due to their low-stakes nature, ILSAs are an ideal ground for innovations and exploring new methodologies. Methods: This study proposes combining state-of-the-art machine translations (i.e., Google Translate & ChatGPT) and artificial neural networks (ANNs) to mitigate two key concerns of human scoring: inconsistency and high expense. We applied AI-based automated scoring to multilingual student responses from eight countries and six different languages, using six constructed response items from TIMSS 2019. Results: Automated scoring displayed comparable performance to human scoring, especially when the ANNs were trained and tested on ChatGPT-translated responses. Furthermore, psychometric characteristics derived from machine scores generally exhibited similarity to those obtained from human scores. These results can be considered as supportive evidence for the validity of automated scoring for survey assessments. Conclusions: This study highlights that automated scoring integrated with the recent machine translation holds great promise for consistent and resource-efficient scoring in ILSAs. [ABSTRACT FROM AUTHOR]

Fuzzy matching techniques are the presently used methods in translating the words. Neural machine translation and statistical machine translation are the methods used in MT. In machine translator tool, the strategy employed for translation needs to handle large amount of datasets and therefore the performance in retrieving correct matching output can be affected. In order to improve the matching score of MT, the advanced techniques can be presented by modifying the existing fuzzy based translator and neural machine translator. The conventional process of modifying architectures and encoding schemes are tedious process. Similarly, the preprocessing of datasets also involves more time consumption and memory utilization. In this article, a new spider web based searching enhanced translation is presented to be employed with the neural machine translator. The proposed scheme enables deep searching of available dataset to detect the accurate matching result. In addition, the quality of translation is improved by presenting an optimal selection scheme for using the sentence matches in source augmentation. The matches retrieved using various matching scores are applied to an optimization algorithm. The source augmentation using optimal retrieved matches increases the translation quality. Further, the selection of optimal match combination helps to reduce time requirement, since it is not necessary to test all retrieved matches in finding target sentence. The performance of translation is validated by measuring the quality of translation using BLEU and METEOR scores. These two scores can be achieved for the TA‐EN language pairs in different configurations of about 92% and 86%, correspondingly. The results are evaluated and compared with other available NMT methods to validate the work. [ABSTRACT FROM AUTHOR]

Recently, generative artificial intelligence (GAI) has impressed the world with its ability to create text, images, and videos. However, there are still areas in which GAI produces undesirable or unintended results due to being "uncertain". Before wider use of AI-generated content, it is important to identify concepts where GAI is uncertain to ensure the usage thereof is ethical and to direct efforts for improvement. This study proposes a general pipeline to automatically quantify uncertainty within GAI. To measure uncertainty, the textual prompt to a text-to-image model is compared to captions supplied by four image-to-text models (GIT, BLIP, BLIP-2, and InstructBLIP). Its evaluation is based on machine translation metrics (BLEU, ROUGE, METEOR, and SPICE) and word embedding's cosine similarity (Word2Vec, GloVe, FastText, DistilRoBERTa, MiniLM-6, and MiniLM-12). The generative AI models performed consistently across the metrics; however, the vector space models yielded the highest average similarity, close to 80%, which suggests more ideal and "certain" results. Suggested future work includes identifying metrics that best align with a human baseline to ensure quality and consideration for more GAI models. The work within can be used to automatically identify concepts in which GAI is "uncertain" to drive research aimed at increasing confidence in these areas. [ABSTRACT FROM AUTHOR]

Context. Today, collecting and creating datasets in various sectors has become increasingly prevalent. Despite this widespread data production, a gap still exists in specialized domains, particularly in the Islamic Friday Sermons (IFS) domain. It is rich with theological, cultural, and linguistic studies that are relevant to Arab and Muslim countries, not just religious discourses. Objective. The goal of this research is to bridge this lack by introducing a comprehensive Sermon Audio and Text (SAT) dataset with its metadata. It seeks to provide an extensive resource for religion, linguistics, and sociology studies. Moreover, it aims to support advancements in Artificial Intelligence (AI), such as Natural Language Processing and Speech Recognition technologies. Method. The development of the SAT dataset was conducted through four distinct phases: planning, creation and processing, measurement, and deployment. The SAT dataset contains a collection of 21,253 audio and corresponding transcript files that were successfully created. Advanced audio processing techniques were used to enhance speech recognition and provide a dataset that is suitable for wide-range use. Results. The fine-tuned SAT dataset achieved a 5.13% Word Error Rate (WER), indicating a significant improvement in accuracy compared to the baseline model of Microsoft Azure Speech. This achievement indicates the dataset’s quality and the employed processing techniques’ effectiveness. In light of this, a novel Closest Matching Phrase (CMP) algorithm was developed to enhance the high confidence of equivalent speech-to-text by adjusting lower ratio phrases. Conclusions. This research contributes significant impact and insight into different studies, such as religion, linguistics, and sociology, providing invaluable insights and resources. In addition, it is demonstrating its potential in Artificial Intelligence (AI) and supporting its applications. In future research, we will focus on enriching this dataset expansion by adding a sign language video corpus, using advanced alignment techniques. It will support ongoing Machine Translation (MT) developments for a broader understanding of Islamic Friday Sermons across different linguistics and cultures. [ABSTRACT FROM AUTHOR]

Regardless of its distinction for touring and tourism communication, the subject of machine translation has not been dealt with regarding tourism investigation. This study aims to bridge this gap in knowledge by exploring machine translation (MT) technology in the development of tourism translation from the perspectives of translators and translation experts in Saudi Arabia. To attain the three objectives of the study, the researcher applied the quantitative approach. For data gathering, a 12-item opinion poll was disseminated among a random sample group that consisted of 80 translators and translation experts who were selected at random from several Saudi universities, top translation services companies, and tourism organizations during the current academic year 1445/AH/2024 CE. The findings of this study discovered that many translators and translation experts had positive perspectives on utilizing machine translation (MT) technology in the development of tourism translation around the world today in general and Saudi Arabia in particular. Moreover, the perspectives of these translators and translation experts on the difficulties and intricacy confronted in applying machine translation (MT) technology in the tourism translation industry were highly positive. Over half of the participants (51.42%) perceived that machine translation (MT) can help overcome obstacles encountered by translators in the field of tourism. The findings revealed that translators and translation experts in the realm of the tourism industry have a promising future thanks to the relationship between machine translation (MT) and tourism translation. [ABSTRACT FROM AUTHOR]

Africa has over 2000 languages; however, those languages are not well represented in the existing natural language processing ecosystem. African languages lack essential digital resources to effectively engage in advancing language technologies. There is a need to generate high-quality natural language processing resources for low-resourced African languages. Obtaining high-quality speech and text data is expensive and tedious because it can involve manual sourcing and verification of data sources. This paper discusses the process taken to curate and annotate text and speech datasets for five East African languages: Luganda, Runyankore-Rukiga, Acholi, Lumasaba, and Swahili. We also present results obtained from baseline models for machine translation, topic modeling and classification, sentiment classification, and automatic speech recognition tasks. Finally, we discuss the experiences, challenges, and lessons learned in creating the text and speech datasets. [ABSTRACT FROM AUTHOR]

Copyright of Journal on Communication / Tongxin Xuebao is the property of Journal on Communications Editorial Office and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Copyright of Journal of Computer Engineering & Applications is the property of Beijing Journal of Computer Engineering & Applications Journal Co Ltd. and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Translation is one of the fields that have benefited from the significant growth that Natural Language Processing has witnessed. ChatGPT is one of the prominent products of such development promoted for its capability of producing human-like translation, among many other things. The main aim of this article is to explore the ability of ChatGPT to translate separable phrasal verbs, which can pose a real challenge to translators as one or more words separate their parts. Therefore, the significance of this paper stems from the pressing necessity to evaluate the utilization of such a tool to address such specific translation issues as it attempts to answer the question of to what extent ChatGPT can translate separable phrasal verbs clearly and accurately. This research paper has followed the qualitative method to analyze the accuracy and clarity of the translation of the separable phrasal verbs generated by ChatGPT. The findings revealed that ChatGPT can translate such phrases clearly, yet human intervention is needed to ensure an accurate delivery. [ABSTRACT FROM AUTHOR]

This study examines the advancements in AI-driven machine translation, specifically focusing on the accurate translation of Arabic colloquial expressions. It aims to assess the progress made by Large Language Models, such as Bing AI Chat, compared to traditional machine translation systems. By focusing on colloquial expressions, this research aims to shed light on the challenges and opportunities for improvement in machine translation systems, particularly when dealing with the complexities of translating informal Arabic utterances. Building upon At-tall's 2019 thesis, which compared Google Translate and human translators, the study employs the same Arabic sentences as a test dataset, allowing for a direct comparison between 2019 translations and those produced by current machine translation tools. The findings indicate limited improvement in Google Translate since 2019, with Bing Translator exhibiting a similar level of translation accuracy. In contrast, Bing AI Chat consistently outperformed the other systems, showcasing the potential of Large Language Model machine translation. Notably, Bing AI Chat provided interpretations and valuable comments on the tested Arabic phrases, demonstrating a deeper understanding of the intended meaning. This study contributes significantly to the field of machine translation by providing evidence of the potential of Large Language Model systems in producing more accurate Arabic-English translations. It emphasizes the advantage of Large Language Models in dealing with non-standard Arabic expressions, encouraging further exploration of Large Language Model-powered approaches in machine translation. The findings offer a promising pathway towards achieving more accurate and expressive translations across diverse languages and cultures. [ABSTRACT FROM AUTHOR]

While the rule of thumb that translators translate into their first language (L1) is often desirable, the demand for translations from lesser used into widely used languages often forces translators into this practice because it outstrips supply. The aim of this study is to show readers how the quality of the reference human translation (REF) produced by translators with Slovak as their L1 can influence the assessment of the quality of machine translation output (MT output), using the example of the light post-edited machine translation output (MTPE output) of technical documentation (TD) from Slovak into German by post-editors (PE) with Slovak as their L1. We therefore use the percentage difference between the TER and HTER scores to compare those segments of the studied TD that the machine translated differently than the human, whereby the PE nonetheless adopted such MT output or only partially changed it. We have divided the results into 6 categories of segments based on the differences between TER and HTER, where the difference between TER and HTER was below 50 %. We found that inappropriate REF and different, but not always wrong, solutions adopted by the PE compared to the REF translator can also result in poor evaluation of the MTPE output. [ABSTRACT FROM AUTHOR]

In recent years, significant progress has been made in the field of non-autoregressive machine translations. However, the accuracy of non-autoregressive models still lags behind their autoregressive counterparts. This discrepancy can be attributed to the abundance of repetitive tokens in the target sequences generated by non-autoregressive models. In this study, we delve into this phenomenon and propose a novel approach to train a non-autoregressive model using unlikelihood loss. We evaluate our method on three widely used benchmark tasks. The experimental results demonstrating that our proposed approach significantly reduces the number of repetitive tokens while improving the overall performance of non-autoregressive machine translations. Compared to the baseline model "Mask-Predict", the average number of repetitions on IWSLT 14 DE → EN valid set is reduced from 0.48 to 0.17, resulting in a remarkable 62% decrease. [ABSTRACT FROM AUTHOR]

In modern times, the number and variety of machine translation technologies are increasing day by day. Currently, there are many translation programs that translate text of any topic from one language to another. The translation made by the programs that compare the various words, sentences, phrases and idioms in their memory during the translation process and find their more appropriate equivalents in the translation language to the source language is considered as a higher quality translation. However, such translation programs are mostly universal programs. In addition to such universal programs, there are also advanced translation programs for translation of some specific field. Each user can choose any among these programs that suits his interests and goals. Of course, all these programs have advantages as well as weaknesses. In our research, the advantages and disadvantages of the most popular translation programs of recent years have been tried to be revealed. A translator who examines these advantages and disadvantages will be able to make a better decision when choosing the translation program he needs. In addition, a detailed analysis of machine translation technologies was made in the research and necessary information about these technologies was given. The research can serve as a database for both translators and researchers in the field of machine translation. [ABSTRACT FROM AUTHOR]

Artificial Intelligence (AI) tools recently became a prominent concern in higher education classrooms. Many teachers have implemented the technology into their assignments, while others are strictly against this technology's use for assignments. Either way, students have found ways to use it in their academic careers. Though research on the power of AI in the workplace exists, research is lacking in its appropriate use in higher education. Universities need to define AI's role on campus and establish guidelines on how these tools may or may not be used and how faculty can recognize misuse, specifically related to academic integrity. This study aimed to determine how faculty view AI as a part of undergraduate literature, language, and linguistics programs. From the interview study, common themes emerged, including implementation, academic integrity, the human aspect of linguistics, and the future of AI writing tools. The faculty interviewed also stated that those in higher education must tread carefully through this strong intersection between technology and the arts to use AI responsibly, strategically, and ethically. [ABSTRACT FROM AUTHOR]

To improve the effect of corpus extraction in bilingual English–Chinese machine translation (ECMT), this article combines a machine learning algorithm with the ECMT corpus extraction method based on bilingual sentence pair coverage. Aiming at the propagation characteristics of medium- and short-range speeches, and employing multi-band speech waveforms in the troposphere, the actual requirements of remote machine translation are combined with the free-space, 1546, and the dual-path models, respectively, to analyze three typical air-to-ground wave propagation models. A simulation is conducted for both predictions and comparisons. In addition, a radio wave propagation model within the line of sight in the troposphere is established. The results suggest that the improved dual-path model is consistent with the propagation characteristics of radio waves in the troposphere. Moreover, the experiments show that the proposed ECMT corpus extraction method based on bilingual sentence pair coverage can play an important role in more accurate English–Chinese translations.

In this article, our aim is to assess the efficacy of machine translation tools and state-of-the-art generative AI platforms such as ChatGPT in fostering enhanced linguistic analysis skills among students in order to improve their understanding of the language. Our case study deals with the analysis of complex English compounds—a known challenge for French learners of English as a second language. Our investigation centers on students’ ability to identify the head noun within complex noun phrases, both in full sentences and in press titles. The study involved two distinct cohorts: students in their third and final year of an undergraduate program in applied foreign languages, and students in their second and final year of a master's program in professional translation. We evaluated the participants’ ability to identify head nouns—a necessary skill for comprehending complex noun phrases—under two conditions: (i) without the aid of any linguistic tools and (ii) with the assistance of machine translation outputs by a generic online translator. Subsequently, we explored the capabilities of advanced generative AI tools—in this instance ChatGPT—of correctly identifying head nouns. Our results show that students may benefit from the presence of machine translation outputs, albeit with varying degrees of success. Our experimentation with ChatGPT shows that, given appropriate prompts, the tool is capable of identifying head nouns, suggesting that generative AI tools may be a more effective tool in helping students analyze and understand complex noun phrases in English.

This article addresses the challenges and opportunities posed by the use of machine translation (MT) and post-editing (PE) in the training of legal translators in the Spanish-Italian combination. The aim is to find out whether students who do not have specific training in MT and PE, but a sufficiently developed translation competence, are able to successfully apply the MT and PE concepts learned during a short workshop in their Specialized Translation course and, as a result, produce a satisfactory post-edited translation of acceptable quality. The results show that students have an inward critical attitude towards the use of MT, so they are not influenced by the MT output. Moreover, they have adequate skills to immediately detect not only most of the errors related to concordance and lack of logic, but also the more hidden ones that require a subsequent revision.

This research is briefly concerned with the field of machine translation of the Noble Qur’an. The main objective of this research is to emphasize the need for the human touch in the translation process, in order to remain faithful to the original semantic, religious, and cultural charge of the Qur’anic text. To measure the accuracy, fluency, and cultural sensitivity of machine translations, we used a mixed approach, combining quantitative content analysis and qualitative thematic analysis. Globally, the results of this research support the use of a hybrid approach combining machine translation with human editing and revision because even though machine translation has improved considerably over the last few decades, it still faces many difficulties in correctly capturing the nuanced meanings, cultural context and the wonderful literary and aesthetic style of the original Arabic-language Qur’anic text. Résumé Cette recherche s’inscrit sommairement dans le domaine de la traduction automatique du saint Coran. L’objectif majeur assigné à cette recherche est d’insister sur la nécessité de l’intervention de la touche humaine dans le processus de traduction afin de rester fidèle à la charge sémantique, religieuse et culturelle originales du texte coranique. Pour mesurer la précision, la fluidité et la sensibilité culturelle des traductions automatiques, nous avons fait appel à une approche mixte, combinant l'analyse quantitative du contenu et l'analyse thématique qualitative. Dans l’ensemble, les résultats de cette recherche favorisent le recours à une approche hybride combinant ainsi la traduction automatique et l’édition et la révision humaine car si la traduction automatique s'est considérablement améliorée au cours des dernières décennies, elle reste confrontée à des difficultés pour saisir correctement les significations nuancées, le contexte culturel et le merveilleux style littéraire et esthétique du texte coranique original en langue arabe.

This paper delves into the potential of machine translation tools, with a specific focus on Google Translate, to expand their role beyond traditional translation tasks to enhance language learning in EFL classrooms. While machine translation tools have become integral to translator training programs, their utilization in language education remains limited and understudied. The present study attempts to highlight the promising avenues for innovative pedagogy in language education by incorporating machine translation tools and EFL students’ attitudes towards them. It explores the application of machine translation in the context of English language acquisition for non-native speakers. The results demonstrate that students consider machine translation as a useful strategy to learn English, and Google Translate (GT) offers advantages in vocabulary expansion and quick translations. However, its limitations, such as reduced accuracy for longer texts and the inability to process idiomatic expressions, necessitate careful consideration when integrating it into language learning curricula. GT can serve as a supplementary tool to support learners, but it should not replace conventional language learning methods. Ultimately, this research emphasizes the need for cautious guidance and monitoring when utilizing automated translation to ensure effective language learning outcomes, bridging the gap between translation and language education while acknowledging the tool's limitations. Résumé Cet article explore le potentiel des outils de traduction automatique, avec un accent particulier sur Google Translate, en élargissant leur rôle au-delà des tâches de traduction traditionnelles pour améliorer l'apprentissage des langues dans les classes EFL. Bien que les outils de traduction automatique soient devenus une partie intégrante des programmes de formation des traducteurs, leur utilisation dans l'enseignement des langues reste limitée et peu étudiée. La présente étude tente de mettre en évidence les pistes prometteuses pour une pédagogie innovante dans l'enseignement des langues en intégrant les outils de traduction automatique et les attitudes des étudiants EFL à leur égard. Cet article explore l'application de la traduction automatique dans le contexte de l'acquisition de la langue anglaise pour les locuteurs non natifs. Les résultats montrent que les étudiants considèrent la traduction automatique comme une stratégie utile pour apprendre l'anglais, et Google Translate (GT) offre des avantages dans l'expansion du vocabulaire et les traductions rapides. Cependant, ses limites, telles que la précision réduite des textes plus longs et l'incapacité à traiter les expressions idiomatiques, nécessitent une attention particulière lors de leur intégration dans les programmes d'apprentissage des langues. GT peut servir d'outil supplémentaire pour soutenir les apprenants, mais ne devrait pas remplacer les méthodes conventionnelles d'apprentissage des langues. Enfin, cette recherche met l'accent sur la nécessité d'une orientation et d'un suivi prudent dans l'utilisation de la traduction automatisée pour assurer des résultats d'apprentissage linguistique efficaces, combler l'écart entre la traduction et l'éducation linguistique tout en reconnaissant les limites de l'outil.

Subtitling has long been the primary type of audiovisual translation in Slovenia. A steady increase in the production of audiovisual content, including English-speaking movies and TV series, along with the development of machine translation, have seen a growing need and opportunity for subtitling globally. The Department of Translation Studies (University of Ljubljana) has offered courses in subtitling to MA students for over twenty years. In the course of this time, teaching methods have been adapted to embrace new technologies. Recently, students have shown interest in incorporating machine translation into the subtitling course. The paper presents an analysis of a subtitling assignment in which students were asked to post-edit DeepL’s Slovene translation of English formatted movie subtitles. The post-edited subtitles were compared with the subtitles done by a group of students from scratch. The results show considerable differences in the students’ multimodal awareness and the overall quality of the subtitles between the two groups.

Artificial Intelligence (AI) applied in the legal field has gained considerable ground in the recent years and is used in many fields, amongst which the legal one is of no exception. This paper wishes to explore the quality of the translation (from English into Italian) of an arbitration clause performed by the ChatGPT chatbot. To do so, the automatically generated target text is post-edited by consulting the BoLC (Bononia Legal Corpus) and the web as corpus in the gov.it Italian governmental domain. General and legal dictionaries are also used. The paper findings report some inaccuracies at word level which are easily tackled by corpus consultation. In view of the results obtained, however, complete reliance upon AI-driven solutions is not recommendable at the time being. Conversely, the use of ad hoc corpora and of targeted web searches are the most feasible and reliable solutions, although time-consuming.

This data article presents a dataset for Siswati, a Bantu language of the Nguni group that is one of the eleven official South African languages and the official language of Eswatini (together with English). The dataset contains parallel textual data between English and Siswati as well as monolingual data for Siswati and was developed for use as training data for machine translation systems, specifically the Autshumato machine translation project. Both corpora can also be used for development and evaluation of Natural Language Processing (NLP) core technologies for Siswati. In addition, the data lends itself for corpus linguistic studies. The article describes how the data was collected, what type of texts it contains and what clean-up was done. It also provides an overview of the number of words contained in the datasets.

ArzEn-MultiGenre is a parallel dataset of Egyptian Arabic song lyrics, novels, and TV show subtitles that are manually translated and aligned with their English counterparts. The dataset contains 25,557 segment pairs that can be used to benchmark new machine translation models, fine-tune large language models in few-shot settings, and adapt commercial machine translation applications such as Google Translate. Additionally, the dataset is a valuable resource for research in various disciplines, including translation studies, cross-linguistic analysis, and lexical semantics. The dataset can also serve pedagogical purposes by training translation students and aid professional translators as a translation memory. The contributions are twofold: first, the dataset features textual genres not found in existing parallel Egyptian Arabic and English datasets, and second, it is a gold-standard dataset that has been translated and aligned by human experts.

Abstract Advances in neural machine translation utilizing pretrained language models (PLMs) have shown promise in improving the translation quality between diverse languages. However, translation from English to languages with complex morphology, such as Arabic, remains challenging. This study investigated the prevailing error patterns of state-of-the-art PLMs when translating from English to Arabic across different text domains. Through empirical analysis using automatic metrics (chrF, BERTScore, COMET) and manual evaluation with the Multidimensional Quality Metrics (MQM) framework, we compared Google Translate and five PLMs (Helsinki, Marefa, Facebook, GPT-3.5-turbo, and GPT-4). Key findings provide valuable insights into current PLM limitations in handling aspects of Arabic grammar and vocabulary while also informing future improvements for advancing English–Arabic machine translation capabilities and accessibility.

Critical thinking and anxiety influenced the translation competence of translators. This study sought to examine the interactions between critical thinking, attitude, and anxiety influenced the translation competence of translators. This study adopted an empirical approach to collect data from 145 student translators from many colleges in Saudi Arabia. The questionnaire was used as a data collection tool. Data were analyzed by using structural equation modelling to find out the relationship between the study factors. The results indicated that there was a negative relationship between AI anxiety with critical thinking and attitude. However, there was a strong positive relationship between attitude with critical thinking, and Machine Translation anxiety. Also, there was a positive relationship between Machine Translation anxiety with AI anxiety and critical thinking.

Natural languages are written and spoken languages, and NLP (Natural Language Processing) is the ability of a computer program to recognize both written and spoken languages. Word Sense Disambiguation (WSD) is identified as a challenging area of research in Artificial Intelligence (AI), and Machine Translation (MT). WSD is the procedure for selecting the exact meaning of a word that has more than one meaning. This is an essential application for all-natural language processing applications. There are various knowledge-based, supervised, and unsupervised approaches to WSD process. The Naïve Bayes classifier as an example of approach supervised and unsupervised approaches is the most important method. In this paper, we emphasize on the use of the Naïve Bayes approach for text classification in WSD techniques. Bayes’ hypothesis is a probabilistic model and a reliable approach for text classification. Bayes’ hypothesis acknowledges that the occurrence of some other features is not dependent on the presence of a particular element in a class. This calculation can be used to solve multi-class prediction problems. This classifier performs better compared to the other methods of different approaches. This paper gives an itemized investigation of Naïve Bayes algorithms, which depicts its ideas, covering up Naïve Bayes,’ text characterization, traditional innocent Bayes,’ and machine learning. We have used the collocation method of feature extraction for the WSD of English sentences. Using this model we have disambiguate ambiguous English words by predicting part-of-speech inclusive of “noun,” “verb,” “adverb,” and “adjective.” This disambiguation module is an enhancement in machine translation. The system reported the performance measure of seventy-eight (78%) percent of the scale on F1-measure.

Paraphrasing, a ubiquitous linguistic practice involving the rephrasing of sentences while preserving their underlying meaning, holds substantial significance across various Natural Language Processing (NLP) applications. This research focuses on the domain of Arabic Paraphrase Generation, aiming to introduce an innovative model capable of generating diverse Arabic paraphrases through experimentation with deep learning model. The proposed model extends beyond conventional baseline approaches, incorporating Transformer-based architectures and ChatGPT models to enhance the richness and variety of generated paraphrases. One notable challenge addressed in this study is the absence of an Arabic parallel paraphrased dataset. Recognizing this gap in existing resources, we propose the creation of an expanded paraphrase corpus, leveraging synthetic artificial data to bolster the paraphrasing generation process. This strategic augmentation aims to not only fill a critical void in the available datasets but also to provide a robust foundation for training and evaluating the performance of the paraphrase generation model. In the experimental phase, various models, including the baseline architecture, and Transformer-based models, are examined to assess their effectiveness in generating meaningful Arabic paraphrases. The results of automatic evaluation reveal that our Fine-tuned GPT-3.5 model surpasses state-of-the-art methods, achieving remarkable scores of 23.69%, 88.30%, and 91.89% in BLEU, BERTScore, and COMET evaluations, respectively. Additionally, the Fine-tuning AraT5v1 model shows around a 2.4% improvement in the BLEU score. Moreover, for human evaluation, Cohen kappa achieved 0.9. These findings highlight the potential of Transformer-based approaches in advancing Arabic Paraphrase Generation and affirm the effectiveness of our proposed model in elevating the quality and diversity of generated paraphrases.

The domain of machine text translation and matching is undergoing substantial transformations amidst the perpetual evolution of deep learning methodologies. By amalgamating the contemporary realm of generative models and networks with the multi-faceted attentiveness of multiple heads, there has been a pronounced enhancement in the efficacy of existing text translation and matching endeavors. Consequently, this manuscript endeavors to elucidate the intricacies of the text-matching conundrum within the ambit of English translation. It posits a novel MA-Transformer text-matching framework that seamlessly integrates multi-tiered semantic feature extraction methodologies to actualize the text-matching task in the English translation process. The framework initiates its journey by employing Continuous Bag of Words (CBOW) for word vector embedding, thereby accomplishing the generation and embedding of word vectors. Subsequently, it expeditiously conducts the multilevel amalgamation of data features through the expeditious execution of the multi-head Transformer model. Following the culmination of feature fusion, a judicious sequence of data downgrading and feature screening ensues, ultimately culminating in the attainment of high-precision text matching. The experimental results show that the constructed MA Transformer model performs well in public and actual data testing, with an average precision of 0.867 and 0.722, respectively, on the two types of datasets. The accuracy of the text-matching is higher than that of the current common method frameworks, which provide technical support and references for the future construction of English translation systems.

Modern text generation metrics use semantic representations of words to assess the quality of a text generation model without considering the fluency of the generated text. This paper proposes a novel text generation metric that combines adequacy and fluency to measure the quality of the generated text. When computing the final score using optimal transport, the metric considers semantic meaning and word order. We evaluate the metric on text translation data sets consisting of 20 language pairs from various language families and scripts. Using a novel statistic for measuring word order sensitivity, we analyze its adequacy-based performance using Pearson’s r and Kendall’s $\tau $ correlation coefficients and their sensitivity to fluency-related modifications. Results show that the proposed metric is the most sensitive to fluency-related changes among all top-performing embedding-based metrics, which were found to be relatively invariant to variations in word order. The proposed metric’s overall adequacy-based performance is lower than the best embedding-based metric but higher than the n-gram matching metrics. Our code is publicly available on GitHub (https://github.com/eriknovak/metric-OPWScore) under the BSD-2-Clause license.

Nowadays, machine translation has been a prevalent Internet application. But there still lacks mature intelligent algorithms to automatically evaluate quality of machine translation results. Considering the complexity inside machine intelligence-based semantic comprehension, we resort to pretraining language model (PLM) to deal with this challenge. Hence, this paper proposes a semantic context context-aware automatic quality scoring method for machine translation based on a specific PLM. The purpose of introducing the calculation of sentiment vectors in research is to consider emotional information in machine translation quality automatic scoring methods, in order to improve the accuracy and robustness of scoring. In particular, a novel PLM that combines multiple key features and tasks is established, which is utilized to make encoding towards largescale initial sentences and object sentences. It is finely tuned by integrating two typical pretraining structures. By applying the proposed PLM to complex semantic context and analysis tasks, we finally demonstrate its effectiveness through experiments on the News Crawl corpus and WMT dataset. The obtained results show that the proposal method has achieved significant improvements in various evaluation indicators, demonstrating its superiority in the quality evaluation of machine translation by perceiving semantic contexts. Through comparison experiments, efficiency of the proposal can be acknowledged.

The field of low-resource sentiment analysis has seen significant developments in recent years. This research review SLR evaluates the approaches and data sources utilized in low-resource sentiment analysis by deep learning. The primary aim is to discover suitable approaches for future sentiment analysis in low-resource. Our studies explore various languages, models, and data sources expressing a desire to create effective approaches. Our emphasis lies in the critical evaluation of the approaches and the datasets utilized, to identify areas where further research is needed. Our analysis study adds to the existing body of literature reviews, encompassing multilingual low-resource sentiment analysis research spanning from 2018 to 2023. The findings indicate that the transfer learning approach is the most frequently used, followed by word embedding learning and machine translation systems. Additionally, the study shows that social media is the most used platform for data collection, followed by product reviews, movies, and hotels. There has been a significant surge in the adoption of pre-trained transformers, indicating a growing interest in exploring the potential of these models for low-resource languages within the natural language processing (NLP) community. This trend is largely attributed to the novel nature of these models and their feature of being non-labour intensive. However, the scarcity of annotated datasets for such languages remains a major hurdle. finally, these research findings are relevant and informative for any researcher working in the field of low-resource multilingual sentiment analysis. The study introduces a conceptual framework for performing sentiment analysis in low-resource. The study provides a valuable resource for future researchers.

In the context of a more linked and globalized society, the significance of proficient cross-cultural communication has been increasing to a position of utmost importance. Language functions as a crucial medium that establishes connections among people, corporations, and countries, demanding the implementation of precise and effective translation systems. This comprehensive review paper aims to contribute to the evolving landscape of AI-driven language translation by critically examining the existing literature, identifying key debates, and uncovering areas of innovation and limitations. The primary objective is to provide a nuanced understanding of the current state of AI-driven language translation, emphasizing the advancements, challenges, and ethical considerations. In this review, ongoing debates surrounding AI-driven language translations were actively involved. By evaluating different viewpoints and methodologies, insights into unresolved questions that contribute to a broader discourse in the field were provided. The future trajectory of this study involves the incorporation of cross-lingual dialect adaptability and the advancement of Artificial Intelligence translation systems, with a focus on prioritizing inclusion and cultural understanding.

Language is a primary means of communication, but it is not the only means; knowing a language does, however, assist speed up the process. Many distinct languages are spoken worldwide, and people use them to communicate. This is only one of the many reasons why language is so crucial. Based on the literature survey, it is evident that there is a lack of available translators for the Tulu language. Despite being prevalent predominantly in Karnataka, the Tulu language has not been as widely spoken as other Indian languages until recently, although it gained enough recognition to become the second language in Karnataka. The purpose of our research work aims at translating the English language into the Tulu language. During the evaluation the system was tested on a dataset consisting of handwritten characters during the evaluation process Convolutional Neural Networks used achieved an accuracy rate of 92%. To translate English to the Tulu language, we employed a parallel sentence dataset for the neural approach and a parallel word dataset for the rule-based approach. The rule-based approach resulted in an 89% accuracy rate for word-based analysis and an 81% accuracy rate for sentence-based analysis for the English-to-Tulu language translation. The neural machine translation approach of the Encoder-Decoder model with LSTM is been used to accomplish translation from English to Tulu with a BLEU score of 0.83 and Tulu to English with a BLUE score of 0.65. The model also employed hybrid machine translation to enhance the translation.

Source code summaries play a crucial role in helping programmers comprehend the behavior of source code functions. In recent deep-learning based approaches for Source Code Summarization, there has been a growing focus on Transformer-based models. These models use self-attention mechanisms to overcome the long-range dependency issue that previous models often encounter, making them a promising solution for the Source Code Summarization task. However, these models suffer from two shortcomings: 1) they are weak in handling the semantics of keywords, and 2) they are weak to learn the source code with complex structure. To resolve these shortcomings, our study proposes integrating Non-Fourier and ASTStructural relative position representations into Transformer-based model for Source Code Summarization, which we have named NFASRPR-TRANS. NFASRPR-TRANS employs two types of positional encoding schemes in two different Transformer encoders. The first encoder handles the semantics of the keywords of the input source code sequence by using the Gaussian Embedder to encode the non-Fourier relative position representation of the sequence. The second encoder uses Tree Positional Encoding to learn the structural information of the Abstract Syntax Trees (ASTs), which provides relative position information in the ASTs for generating the source code summaries. Finally,we compared NFASRPR-TRANS with previous models and evaluated its performance on the Java and Python datasets using five metrics, including BLEU, ROUGE-L, CIDEr, METEOR, and SPICE. NFASRPR-TRANS achieves 2%-10% improvements across all five metrics on both datasets.

We introduce BERT-NAR-BERT (BnB) – a pre-trained non-autoregressive sequence-to-sequence model, which employs BERT as the backbone for the encoder and decoder for natural language understanding and generation tasks. During the pre-training and fine-tuning with BERT-NAR-BERT, two challenging aspects are considered by adopting the length classification and connectionist temporal classification models to control the output length of BnB. We evaluate it using a standard natural language understanding benchmark GLUE and three generation tasks – abstractive summarization, question generation, and machine translation. Our results show substantial improvements in inference speed (on average 10x faster) with only little deficiency in output quality when compared to our direct autoregressive baseline BERT2BERT model. Our code is publicly released on GitHub (https://github.com/aistairc/BERT-NAR-BERT) under the Apache 2.0 License.

In a rapidly evolving language services landscape where the boundaries between translation, revision, and post-editing are becoming increasingly blurred, this article provides a state-of-the-art review on the intricate relationship between human translators and Neural Machine Translation (NMT). While acknowledging the significant advancements made in the field of Artificial Intelligence (AI) applied to automatic language processing, we want to emphasize the pressing need to reassert the value of human translators over unconscious algorithms that automatically generate translations. This paper also highlights the synergy between human skills and technology, positioning NMT as a key additional tool in a translator’s toolkit that enhances both speed and efficiency to meet the exponentially growing demand for translation services and to cope with the ever-tightening deadlines. Nevertheless, considering the ambivalent effects of NMT on the quality of the final product, we advocate for the informed and responsible use of machine translation (MT) tools by both professionals and translation students. Furthermore, we believe that beyond fostering MT literacy, translation trainers may play a crucial role in nurturing adaptability in future translators, equipping them with the necessary skills and tools to confidently navigate the latest and emerging technologies—such as interactive and adaptive MT tools or chatbots—in order to secure them a competitive advantage in the translation industry.

Background and Objectives: Although Machine Translation (MT) is extensively researched within the field of Artificial Intelligence (AI) and translation studies, few studies have attempted to implement MT output in foreign language teaching (FLT). One potential aspect of using MT in FLT refers to the implementation of MT output for reading comprehension. Considering the existing gap in the body of research on this issue, the present study aimed to investigate whether MT output is qualified enough to be used as an aid in EAP reading comprehension courses. More specifically, this study intended to benchmark the efficacy of MT output for EAP reading comprehension courses based on the data obtained from testing its comprehensibility and probing the students’ perceptions. To achieve the objectives of the study, MT was operationally defined as quality assessment in terms of output efficacy, a combination of usability and comprehensibility, which mirrors the ultimate goal of MT use in EAP reading comprehension courses, from the users' or target readers’ standpoint. Within this perspective, the current research was an attempt to assess the quality of MT output in terms of comprehensibility and the degree to which MT output might be comprehensible to the EAP students participating in this study.Materials and Methods: The participants of the study, 140 Iranian undergraduate university students majoring in the field of education at Farhangian University, Iran, were selected based on simple random sampling. Oxford Quick Placement Test was used to homogenize them in terms of English proficiency. Two versions of a reliable reading comprehension test, human translation (HT) and Machine Translation (MT), were given to. This test included 25 multiple-choice items, assessing the participants' literal comprehension of information stated in the passage as well as higher-order comprehension that required making inferences and conclusions. In particular, the items measured textual coherence, inference, reference, scanning, skimming, and word-meaning inference. To test the reliability of the tests, the KR-21 formula was applied and the results showed that both HT test (.83) and MT test (.78) were reliable. To investigate the perceptions of the participants on the efficacy of the MT output they encountered on the test, semi-structured interviews were conducted with some of the participants in Persian. Findings: With reference to the results of non-parametric tests such as Spearman’s rho, and Mann-Whitney Tests, and considering the observed effect sizes (Cohen’s d), it was revealed that, generally, the efficacy of MT output is comparable to that of HT. Moreover, in terms of reading comprehension sub-skills, the qualities of the two translations were comparable with regard to scanning, and inference, but not skimming and reference. Furthermore, the findings from the interview indicated that the students perceive MT to be a seminal aid for their EAP reading comprehension activities despite the minor problems that exist in the output such as morpho-syntactic errors or inappropriate lexical equivalents.Conclusions: The present study confirmed the fact that the efficacy of MT output is target-reader-dependent and text-dependent since it is determined both by the characteristics of the readers, such as their disciplines, and text features, as demonstrated by the significant differences in comprehension levels of the same readers measured by the same questions for HT and MT output. Accordingly, this study shed limelight on comprehensibility as a criterion of MT output efficacy; that is to say, it has to be reminded that MT quality needs to be defined as a context-bound and target-reader-specific concept.

The purpose of this study is to evaluate ChatGPT-4’s language proficiency in Italian dialects. At the outset, it is clarified what is meant by ‘language ability’ within the context of Large Language Models. This involves identifying the tasks that ChatGPT might face in real-world scenarios, from which we can derive inferential assumptions regarding its linguistic ability. The skills identified, which served as foundation for test design, include comprehension and translation, dialect recognition, analysis of the distinctive features, error detection, text production, interaction, theoretical background, and self-assessment. The tests were crafted to mimic situations requiring these competencies, trying to emulate authentic ChatGPT-User interactions. The results highlight ChatGPT’s excellent prowess in understanding and recognizing Italian dialects and their subvarieties, and a robust background and awareness of its own knowledge. However, the model exhibits significant gaps in analytical skills and struggles with text production and interactive tasks, suggesting superior passive linguistic capabilities compared to active ones.

Technology in the context of education is a subject of debate, from a very positive experience that promotes learning to a very negative one that prohibits the use of various smart devices, tools, programmes, and platforms in the classroom. The problem is how to find a balance between the two positions and how to encourage teachers to introduce possibilities of technologies to benefit the general education process. The topic of machine translation in educational contexts has gained the attention of the research community only recently. Previous studies not only point to the benefits that the technology may bring to the classroom, especially in foreign language learning, but also report mixed views of educators. This study, which is based on the findings from a survey of Lithuanian secondary school teachers, seeks to explore the current status of the inclusion of machine translation in the educational process from teachers’ perspective to envisage the teacher’s role as a facilitator or a mediator in developing children’s machine translation literacy. The conclusions that can be drawn imply that machine translation is rarely considered to be a useful technology by teachers, and its benefits are either unknown or underestimated. Therefore, the need for machine translation literacy instruction emerges.