Measuring the Technique and Quality of Instagram Caption Auto-Translation to Improve @alarabiya Account Translation Results

In this era of globalisation, translation technologies have become more popular in daily communication, the education sector, and the translation industry. It is observed that there is a prevalent use of machine translation (MT) among translation learners. The proper use versus abuse of MT can be a critical issue regarding its role in and impact on translation teaching. This exploratory study aims at investigating learners’ and instructors’ knowledge of MT, experience in MT use, perceived MT quality, ethics of MT use, and the perceived relationship between MT and translation training, in order to figure out the usefulness of MT in translation competence acquisition and the necessity of MT training. To this end, we conducted surveys and semi-structured interviews and found that the influence of MT in translation competence acquisition is determined by the properties of MT and learners’ quality. MT is particularly helpful in gaining lexical knowledge and knowledge to ensure translation efficiency, but not in bicultural knowledge. However, such usefulness builds on learners’ language proficiency, analytic ability, and learning motivation. In light of the findings, issues including the sustainability of MT from ethical and linguistic perspectives, and the potential and proper use of MT to inform translator training, are discussed.

While the quality of machine translation is getting better and better, it is still not perfect. In this case, how the user treats the imperfect machine translation results determines whether the user achieves the purpose for using machine translation. In this process, user evaluations are a key factor. User evaluations are not objective nor accurate. User evaluations are a subjective evaluation of the user and is related to the user. Therefore, this paper attempts to analyze the relationship between the level of the user's L2 and user evaluations. This paper surveyed 69 Chinese users to understand the current status of their use of machine translation, including frequency of use, purpose of use, favorite machine translation, reasons for preferences, and satisfaction with the quality of the machine translation used. At the same time, for specific translation articles, it lets users evaluate the accuracy and fluency of machine translation articles. The survey results show that users with low L2 levels used machine translation at a higher frequency and were more inclined to evaluate machine translation from the perspective of ease of use. In addition, statistical analysis of the evaluation results found that users’ evaluation of machine translation was related to the user's own level L2. The lower the L2 level, the higher the evaluation of the adequacy and fluency of machine translation, and the higher the assessment of the overall quality of the machine translation.

Given the rich body of technical developments and a relatively long history of industrial use of Machine Translation (MT), it is astonishing of how little interest the topic of MT quality has received so far. In this paper, we present three ways of performing MT quality evaluation from our own research: (1) using TQ-AutoTest, a framework for semi-automatic testing and comparison of different translation engines; (2) applying the multidimensional quality metrics for analytical markup of translation errors; and (3) performing task-based user testing. We will set these three in perspective as they serve different needs and different people’s interest in translation quality assessment. This paper deals with the translation of text in the first place. Still, we hope that the methods, insights, and observations we report transfer to broader applications of translation in the field of Media Accessibility.

The use of machine translation has become ubiquitous across various translation practices, especially with the advent of neural machine translation and the integration of deep learning and artificial intelligence in translation program development. While the accuracy and quality of machine translation outcomes have significantly improved, challenges persist particularly in legal translation from English to Arabic. The unique nature of legal discourse and structural differences between English and Arabic make accurately translating legal language features a daunting task. This study aims to evaluate the quality of neural machine translation in rendering legal Latin phraseology into Arabic by comparing two websites: Google Translate and Yandex. A corpus-based approach was adopted where 270 Latin-origin legal terms and phrases were collected, scrutinised, and translated using both platforms. The evaluation focuses on four criteria: inappropriate translations, no translations provided, borrowing (phonetic transliteration into Arabic), and equivalence—the culturally and functionally suitable translation. Key findings indicate that despite significant advancements in machine translation technology, accuracy remains a critical issue, with approximately half of the terms not translated correctly. While Google Translate is widely used, Yandex demonstrated higher accuracy in this context. Furthermore, the majority of phrases selected for this study were not accurately translated by either website. The solution to this problem lies in enhancing the training process. Arabic users and translators should contribute more translations to enrich Arabic corpora online. Additionally, it's been observed that there is a lack of English-Arabic dictionaries or databases dedicated to Legal Language Processing (LLP). Therefore, initiating a research project addressing this issue could be of utmost importance. Regarding specialized language, improving the quality of Neural Machine Translation (NMT) raises questions about its reliability for both learners and professional translators. Accordingly, the study recommends further research on assessing machine translation quality, improving neural machine translation terminology accuracy, and enhancing machine learning models with more Arabic content and corpora.

To investigate L2 adolescent learners’ use of machine translation (MT), an MT error correction (EC) test was developed, based on the analysis of MT errors arising from translating the learners’ L1 of middle school EFL textbooks. Learners were also asked to report on their use of MT EC strategies on the EC task. Results indicated that mistranslated sentence and verb tense are the most difficult types of MT errors to correct. Furthermore, to resolve MT errors, guessing from context and literal translations were the two most frequently employed EC strategies. When multiple regression analysis was conducted to examine the contribution of EC strategies to the learners’ ability to correct errors, the mid proficiency learners’ reliance on literal translations and the low proficiency learners’ use of multiple EC strategies were positively associated with improved corrections of MT errors. The results of the study are discussed in light of how L2 learners need to develop competence for using MT in L2 writing.

This study explores the interaction effect between source text (ST) complexity and machine translation (MT) quality on the task difficulty of neural machine translation (NMT) post-editing from English to Chinese. When investigating human effort exerted in post-editing, existing studies have seldom taken both ST complexity and MT quality levels into account, and have mainly focused on MT systems used before the emergence of NMT. Drawing on process and product data of post-editing from 60 trainee translators, this study adopted a multi-method approach to measure post-editing task difficulty, including eye-tracking, keystroke logging, quality evaluation, subjective rating, and retrospective written protocols. The results show that: 1) ST complexity and MT quality present a significant interaction effect on task difficulty of NMT post-editing; 2) ST complexity level has a positive impact on post-editing low-quality NMT (i.e., post-editing task becomes less difficult when ST complexity decreases); while for post-editing high-quality NMT, it has a positive impact only on the subjective ratings received from participants; and 3) NMT quality has a negative impact on its post-editing task difficulty (i.e., the post-editing task becomes less difficult when MT quality goes higher), and this impact becomes stronger when ST complexity increases. This paper concludes that both ST complexity and MT quality should be considered when testing post-editing difficulty, designing tasks for post-editor training, and setting fair post-editing pricing schemes.

This monograph offers a comprehensive analysis of machine translation (MT) acceptance, awareness, and quality within Lithuanian society. It moves beyond the predominantly professional-centric perspectives found in existing research to explore the multifaceted experiences of a wide range of users, encompassing students, professionals, and the general public. The study utilizes a mixed-methods approach, combining quantitative survey data with qualitative insights gathered through interviews and eye-tracking experiments. This methodology allows for a nuanced understanding of how various social groups perceive, use, and evaluate MT technologies, considering factors such as usability, satisfaction, and the recognition of potential risks. The research investigates the acceptability of machine-translated output, exploring the degree to which users find it satisfactory and appropriate for their needs. It also examines public awareness of MT capabilities and limitations, focusing on the level of understanding among different user groups regarding the technology’s strengths and weaknesses. A significant aspect of the study is the assessment of machine translation quality, considering both the linguistic accuracy and the overall effectiveness of the output. The findings shed light on the interplay between user expectations, technological capabilities, and societal contexts in shaping the use and acceptance of MT. The monograph is organized into six chapters, beginning with an overview of theoretical and methodological frameworks. Subsequent chapters delve into empirical findings, revealing detailed analyses of user experiences, opinions, and behaviors. Key themes explored include the impact of age, education level, and professional background on MT use and perceptions; the prevalence and nature of MT applications along with users’ strategies for addressing limitations in MT quality. The conclusions synthesize the research findings, highlighting both the opportunities and the challenges posed by MT in a low-resource language setting like Lithuanian, offering valuable insights for researchers, policymakers, and practitioners in the translation and technology fields.

BackgroundChinese is the second most common language spoken by limited English proficiency individuals in the United States, yet there are few public health materials available in Chinese. Previous studies have indicated that use of machine translation plus postediting by bilingual translators generated quality translations in a lower time and at a lower cost than human translations.ObjectiveThe purpose of this study was to investigate the feasibility of using machine translation (MT) tools (eg, Google Translate) followed by human postediting (PE) to produce quality Chinese translations of public health materials.MethodsFrom state and national public health websites, we collected 60 health promotion documents that had been translated from English to Chinese through human translation. The English version of the documents were then translated to Chinese using Google Translate. The MTs were analyzed for translation errors. A subset of the MT documents was postedited by native Chinese speakers with health backgrounds. Postediting time was measured. Postedited versions were then blindly compared against human translations by bilingual native Chinese quality raters.ResultsThe most common machine translation errors were errors of word sense (40%) and word order (22%). Posteditors corrected the MTs at a rate of approximately 41 characters per minute. Raters, blinded to the source of translation, consistently selected the human translation over the MT+PE. Initial investigation to determine the reasons for the lower quality of MT+PE indicate that poor MT quality, lack of posteditor expertise, and insufficient posteditor instructions can be barriers to producing quality Chinese translations.ConclusionsOur results revealed problems with using MT tools plus human postediting for translating public health materials from English to Chinese. Additional work is needed to improve MT and to carefully design postediting processes before the MT+PE approach can be used routinely in public health practice for a variety of language pairs.

Currently, information technology is rapidly developing and one of its branches can be called machine translation. The use of machine translation in the process of understanding each other by people from different countries is increasing every year. At the moment, Google and Yandex machine translations are among the best machine translations. The quality of machine translation from Yandex and Google is improving every year. However, according to the results of the experiment, when translating from English or Russian into Kazakh and Turkic languages, the quality of the translation decreases. This was shown by the translation result obtained from these two machine translations in March 2024. After all, translation has also shown that it is directly related to the structure of language. Since 2000, scientists from the state of Kazakhstan have been actively studying translations into the Kazakh language. The goal of the work is to improve the quality of translation from English into Kazakh. For this purpose, a transforming model was created for the Kazakh and Turkic languages for learning translation in neural machine translation OpenNMT(). The created model studied and learned an English-Kazakh parallel corpus of 180,000 words. Later, the document with a structure of 20,000 different English sentences was translated into Kazakh. The result is measured using the Blue() metric. The translation result showed a high level. It is shown that in order to improve the results of the experiment carried out in the work during model training, it is necessary to increase the number of parallel corpora created from the English-Kazakh language pair.

Currently, information technology is rapidly developing and one of its branches can be called machine translation. The use of machine translation in the process of understanding each other by people from different countries is increasing every year. At the moment, Google and Yandex machine translations are among the best machine translations. The quality of machine translation from Yandex and Google is improving every year. However, according to the results of the experiment, when translating from English or Russian into Kazakh and Turkic languages, the quality of the translation decreases. This was shown by the translation result obtained from these two machine translations in March 2024. After all, translation has also shown that it is directly related to the structure of language. Since 2000, scientists from the state of Kazakhstan have been actively studying translations into the Kazakh language. The goal of the work is to improve the quality of translation from English into Kazakh. For this purpose, a transforming model was created for the Kazakh and Turkic languages for learning translation in neural machine translation OpenNMT(). The created model studied and learned an English-Kazakh parallel corpus of 180,000 words. Later, the document with a structure of 20,000 different English sentences was translated into Kazakh. The result is measured using the Blue() metric. The translation result showed a high level. It is shown that in order to improve the results of the experiment carried out in the work during model training, it is necessary to increase the number of parallel corpora created from the English-Kazakh language pair.

Clear translation remains a major challenge to better communication and understanding of the international academic literature, despite advances in Machine Translation (MT). Automatic translation systems which captured the detail and the sense of any manuscript in any language for a reader from any other linguistic background would find global applications.In this article, we discuss the current opportunities and constraints to the wider use of machine translation and computer-assisted human translation (CAT). At the present stage of technology development, these instruments offer a number of advantages to specialists working with scientific texts. These include the facility to skim and scan large amounts of information in foreign languages, and to act as digital dictionaries, thesauri and encyclopedias. Word-to-word and phrase-to-phrase translation between many languages and scripts is now well advanced.The availability of modern machine translation has therefore changed the work of specialist scientific translators, placing greater emphasis on more advanced text and sense editing skills. However, machine translation is still challenged by the nuances of language and culture from one society to another, particularly in the freestyle literature of the arts and humanities. Scientific papers are generally much more structured, but the quality of machine translation still largely depends on the quality of the source text. This varies considerably between different scientific disciplines and from one author to another.The most advanced translation systems are making steady progress. It is timely to revisit traditional training programmes in the field of written translation to focus on the development of higher-level research competencies, such as terminology search, and so to make best use of evolving machine translation technologies.More widely, we consider that there is a challenge across the higher education systems in all countries to develop a simple, clear and consistent “international” writing style to assist fast, reliable and low-cost machine translation and hence to advance mutual understanding across the global scientific literature.

Regardless of recent arguments about the wide-scale capabilities of artificial intelligence introduced into machine translation systems, some professionals still underestimate machine translation. However, other scholars see MT as an opportunity to develop and improve the translation industry. Apparently, there is no doubt that machine translation has massively impacted the translation profession and changed radically the way human beings interact through languages. Therefore, translators, university professors, and translation companies seek to adapt to these radical transformations in the field of translation studies. Regardless of the advantages of machine translation, it still confronts huge challenges especially when MT strategies and procedures are applied to specific texts in different contexts particularly colloquial Arabic extensively proliferated in contemporary literature and mass media. On this basis and in response to repeated claims about the high efficiency of machine translation and its extra-ordinary potentialities to render any text from one language into another with accuracy and precision, this paper emphasizes the inability of machine translation mechanisms to render Arabic texts into English and vice versa. The paper emphasizes the damaging impact of using machine translation in rendering into English not only colloquial Arabic dialects but also modern standard Arabic (MSA). The paper also underlines the translation errors resulting from the use of machine translation in rendering modern English texts into Arabic and vice versa with focus on the translation of idiomatic expressions and proverbs. As applied study, the paper will use a variety of texts selected from various literary and non-literary sources/contexts and translate them by “Google Translate” to underline the drawbacks of MT, which subsequently lead to the distortion of the meaning of the SL texts translated into TL. In other words, the paper aims to uncover the mistakes resulting from the use of machine translation when converting both MSA and colloquial Arabic expressions into English and vice versa. The argument of the paper consists of four parts including an introduction, which navigates contemporary translation theories, followed by a scrutiny of the challenges confronting Arabic-English translation, and examples of Arabic/English/Arabic carried out by machine and human translation in addition to a conclusion.

Rapid improvements in the capabilities of machine translation (MT) raise questions about possible increases in overreliance on MT among lower‐proficiency or novice level language learners. This study investigated how such learners described their use of online MT for independent reading and writing tasks, and whether this included descriptions of second language (L2) avoidance behavior. We also explored learners' reasons for using MT and the perceived effects on their language learning. Findings from in‐depth interviews with eight second‐year tertiary language learners suggest that using MT could exceed desirable use among such learners in relation to the language learning objectives, resulting in language avoidance. Although MT helped them in completing language tasks, its effects were perceived to be detrimental toward their abilities to express themselves in the L2. As such, the use of MT may lead to purely superficial language learning in formal language programs. These findings suggest language educators need to consider instructional scaffolding in language programs for such learners and guidelines to assist their autonomous use of the tool.

The pros and cons of which one is better in producing good result of translation between Machine Translation (MT) and Human Translation (HT) has been going on for many years. In the attempt to observe which is better between MT and HT, this article focuses on exploring the techniques used by U-Dictionary as a MT and Maggie Tiojakin as a HT in translating The Gift of the Magi into Indonesian. Data in this research are the words, phrases, clauses and sentences related to the translation techniques in the original version of The Gift of the Magi and the two translation versions. The collected data are analyzed qualitatively by using Molina and Albir’s (2002) theory. The results describe that Maggie Tiojakin used 12 techniques; they are adaptation, amplification, compensation, description, discursive creation, established equivalent, generalization, literal translation, modulation, particularization, reduction, and transposition. Meanwhile, U-Dictionary used 8 techniques; they are amplification, borrowing, calque, established equivalent, literal translation, modulation, reduction, and transposition. The dominant translation technique used by Maggie Tiojakin is discursive creation (24.54%), whereas in U-Dictionary, it is literal translation (47.27%). From the different translation techniques used, it can be proven that HT uses more various techniques and has better translation result than MT, in which the translation of the literary works especially a short story done by HT is more accurate, readable, and acceptable.

The advent of social media has shaken the very foundations of how we share information, with Twitter, Facebook, and Linkedin among many well-known social networking platforms that facilitate information generation and distribution. However, the maximum 140-character restriction in Twitter encourages users to (sometimes deliberately) write somewhat informally in most cases. As a result, machine translation (MT) of user-generated content (UGC) becomes much more difficult for such noisy texts. In addition to translation quality being affected, this phenomenon may also negatively impact sentiment preservation in the translation process. That is, a sentence with positive sentiment in the source language may be translated into a sentence with negative or neutral sentiment in the target language. In this paper, we analyse both sentiment preservation and MT quality per se in the context of UGC, focusing especially on whether sentiment classification helps improve sentiment preservation in MT of UGC. We build four different experimental setups for tweet translation (i) using a single MT model trained on the whole Twitter parallel corpus, (ii) using multiple MT models based on sentiment classification, (iii) using MT models including additional out-of-domain data, and (iv) adding MT models based on the phrase-table fill-up method to accompany the sentiment translation models with an aim of improving MT quality and at the same time maintaining sentiment polarity preservation. Our empirical evaluation shows that despite a slight deterioration in MT quality, our system significantly outperforms the Baseline MT system (without using sentiment classification) in terms of sentiment preservation. We also demonstrate that using an MT engine that conveys a sentiment different from that of the UGC can even worsen both the translation quality and sentiment preservation.