Direction matters: Comparing post-editing and human translation effort and quality.

This experimental study aims to evaluate the quality of post-edited texts, originally translated using computer-assisted translation (CAT) tools, in comparison with traditional human translation. This study investigates the quality of post-editing (PE) compared to traditional translation from scratch (TFS) in the context of Arabic–English translation, utilizing the Phrase CAT tool. The main hypothesis posits that PE yields a final product whose quality is similar or equivalent to that of TFS. The participants’ scores and error frequencies were evaluated using the American Translators Association framework for standardized error marking, and terminology, word choice, mistranslation, addition/omission, spelling, punctuation, case, inconsistency, style, and grammar in both approaches were compared. Data from nine professional Saudi translators showed that PE generally outperformed TFS in terminology, spelling, punctuation, and case, whereas TFS exhibited strengths in consistency, style, grammar, and literal translation. Statistical analysis confirmed the similarity in overall error rates between PE and TFS. The difference in mean error numbers between TFS and PE was not statistically significant. Thus, the disparity in means likely resulted from random chance and might not indicate substantive differences between the two groups. These results imply that PE yields quality that is comparable or equivalent to that of TFS, proving the aforementioned hypothesis. The implications highlight the need for CAT tool training and PE skills among translators to meet the demands of evolving translation technologies. Furthermore, this study underscores the importance of integrating PE training into translation curricula and organizing workshops to improve CAT tool usage.

This study aims to compare and analyze the translation effects of low-resource languages in Asia based on the Pivot Language method. By selecting five different low-resource languages in Asia (Urdu, Burmese, Kyrgyz, Kazakh, and Tajik), this study explores the translation effect of using a pivot language (usually English) to build a bridge between the source language and the target language, and compares it with direct translation. The study uses two translation modes, Pivot Translation Technique and Pivot Prompting, to compare the quality differences between translation through a pivot language and translation directly from the source language to the target language. By calculating translation quality evaluation indicators such as Comet and BLEU, this study evaluates the performance of the two translation modes in different languagesand analyzes whether adding a pivot language can effectively improve translation quality, especially in terms of grammatical accuracy, semantic transmission, and fluency of expression. The results show that the pivot language method significantly improves the accuracy and fluency of translation in low-resource language translation, especially in the translation of long and difficult sentences, and the pivot language translation has higher quality than direct translation. This discovery provides a new direction for machine translation research in low-resource languages and is of great significance to the country's cross-language communication and cultural cooperation.

This paper reports on the results of a process-oriented longitudinal study focusing on the use of online translation resources by translation trainees during a specific human translation (HT) or post-editing (PE) task. It focuses on how students’ resource dominance changes over the four semesters of training: human translators’ and post-editors’ online resource use is also compared, as are students’ research strategies in terms of (1) the types of resources, (2) the frequency of resource use and (3) the time spent on resource consultation at the beginning and the end of their training.

With the rapid advancement of technology, machine translation has been increasingly applied in the field of medical translation. As the foundational text of Traditional Chinese Medicine (TCM), the Huangdi Neijing plays a critical role in shaping TCM theory and clinical practice. This study focuses on the translation of four-character terms in the Huangdi Neijing and compares the translation quality between human and machine translation. A mixed-methods approach, incorporating both qualitative and quantitative analysis, was employed. A total of 463 samples were selected, and four mainstream machine translation systems—Youdao, DeepL, ChatGPT-4o, and DeepSeek were utilized.By analyzing BLEU and TER scores to evaluate the outputs against human translations, it was found that ChatGPT-4o achieved the highest BLEU score of 0.60, indicating the greatest lexical similarity to human translation, and the lowest TER score of 0.99, suggesting the fewest required edits. These results suggest that ChatGPT-4o is currently the most effective among the tested systems in handling this specialized translation task. Despite these advancements, machine translation still exhibits notable limitations when applied to TCM texts. Human translation still outperforms machine-generated outputs, particularly in conveying semantic precision and preserving the cultural and conceptual subtleties embedded in medical discourse. This study contributes to the field by providing empirical evidence on the performance of large language models in translating culturally and linguistically complex four-character expressions from classical TCM texts.

The aim of this article is to investigate translation trainees’ perceived difficulties and benefits of a post-editing oriented neural machine translation (NMT) error annotation and quality evaluation task which was carried out for the language pair English-Turkish and for two separate text types and domains, i.e., environmental blogs and movie/TV reviews, within the scope of an MA course on translation quality standards. The data to be analyzed were collected through a semi-structured questionnaire which was given to the trainees attending the course after the completion of the task. The questionnaire was prepared with the aim of understanding perceived difficulties and benefits of the task. Analysis of the answers revealed that most of the trainees were in the opinion that the task was difficult. Majority of the trainees also believed that the task was beneficial and enabled them to feel empowered to make decisions on human translation quality evaluation of machine translation (MT) and to carry out error annotation and post-editing activities in the future. According to a significant number of trainees, error annotation facilitated their post-editing process and reduced the effort in post-editing. Enhanced understanding of MT error annotation and enhanced ability to perform post-editing were the significant benefits stated by the trainees. The difficulties were associated with being introduced and assigned to perform tasks they were not familiar with. Yet, as displayed in the answers to the questionnaire, a considerable majority of the trainees were positive about the learning experience. The results have also shown that integration of MT-related activities into translator training with a focus on the empowerment of the human translator has its difficulties and benefits also for translation trainers. While the difficulties for the trainers concern the decisions on the design, implementation, and planning of the task and the responsibility to carry out the task in constant interaction and collaboration with the trainees, the benefits are the sense of fulfillment and enrichment brought by positive feedback from the trainees and the discovery of the fact that the so-called ‘teaching experience’ becomes a ‘learning experience’ for trainers as well as trainees.

The progress of Artificial Intelligence brings about profound developments in Computer Aided Translation (CAT). Therefore, many language service providers use CAT tools to improve translation efficiency while ensuring quality. With people's attention to health deepening, popular medical texts are becoming mainstream. This article studies the translation of medical texts in the mode of Machine Translation and Post-editing (MTPE). The source text This Is Your Brain on Food is pre-translated by Google Translate and post-edited by the translator. Based on this practice, the Machine Translation (MT) errors are summarized, and Post-editing (PE) solutions are illustrated at the lexical level. There are MT errors about mistranslation and omission. Several PE solutions are presented in the article, which cover replacement, supplementation and conversion of part of speech. It is shown that the MTPE mode can improve the efficiency of medical translation. Post-editing can ensure the translation effect as equal as possible while faithfully conveying the original information, so as to improve the quality of the target text and assist medical development through translation.

Compared to fully manual translation, post-editing (PE) machine translation (MT) output can save time and reduce errors. Automatic word-level quality estimation (QE) aims to predict the correctness of words in MT output and holds great promise to aid PE by flagging problematic output. Quality of QE is crucial, as incorrect QE might lead to translators missing errors or wasting time on already correct MT output. Achieving accurate automatic word-level QE is very hard, and it is currently not known (i) at what quality threshold QE is actually beginning to be useful for human PE, and (ii), how to best present word-level QE information to translators. In particular, should word-level QE visualization indicate uncertainty of the QE model or not? In this paper, we address both research questions with real and simulated word-level QE, visualizations, and user studies, where time, subjective ratings, and quality of the final translations are assessed. Results show that current word-level QE models are not yet good enough to support PE. Instead, quality levels of > 80% F1 are required. For helpful quality levels, a visualization reflecting the uncertainty of the QE model is preferred. Our analysis further shows that speed gains achieved through QE are not merely a result of blindly trusting the QE system, but that the quality of the final translations also improves. The threshold results from the paper establish a quality goal for future word-level QE research.

The translation of intangible cultural heritage (ICH) texts presents significant challenges due to their rich cultural connotations and unique regional characteristics. To address this, the author of this study utilized Corpus and Computer-assisted Translation (CCAT) to assist in the English translation of ICH during the creation of a parallel corpus. Three patterns of CCAT-based English translation of ICH were proposed: machine translation, human-machine interactive translation (combining human and machine translation), and human translation. The implementation of these patterns can improve the efficiency and quality of ICH translation, promote terminology standardization, and facilitate the exportation and transmission of ICH.

Recently, the improvement of artificial intelligence network translators has emerged as a social issue, and vague anxiety and expectations are coexisting due to AI that goes far beyond human capabilities. Research on the accuracy and quality of machine translation is underway due to the increase in learners learning foreign languages using artificial intelligence machine translation. Existing research focused on error analysis, but this study was conducted as a post-editing activity that learners correct and improve while comparing the results of machine translation of Google and Papago with human translation. This study focuses on how translation education should be changed as a foreign language learning, and look at the direction of AI machine translation and foreign language education, such as whether post-editing education is possible as a foreign language learning and whether it helps improve foreign language skills.

Post-editing (PE) machine translations (MT) has become an increasingly common practice in the translation field in recent years. Research has investigated, among other issues, the types of error corrected by post-editors, but less emphasis has been placed on the corrections themselves and how they reflect MT errors. This article presents a pilot study analysing the edits made by five student post-editors in an English–Finnish post-editing task. We analyse the correctness and necessity of the edits. Our results show that, whereas most edits performed in the task are correct, a significant number of them (34%) are unnecessary. The findings suggest that specific types of edit, such as word-order changes and deletions of personal pronouns, are generally unnecessary for this language pair, which may have implications for post-editing practice and training.

The past decades have witnessed remarkable progress in machine translation (MT) quality, sparking a heated debate within and beyond academia about whether human translators will be replaced by MT systems. This case study conducts a genre-based manual quality evaluation of Chinese-to-English translations produced by Google Translate, Baidu Translate, Sogou Translate, Youdao Translate, and human translators. The findings reveal that the quality of online machine translation output remains incomparable to that of human translation across expressive, informative, and vocative genres. Sogou Translate excels in expressive and informative genres but requires significant improvements in fidelity and comprehensibility for expressive genres, as well as in fidelity and genre function reproducibility for informative genres. For vocative genres, the quality of online machine translation varies across the four evaluation criteria and necessitates targeted enhancements. Notably, Google Translate, despite its reputation as a leading machine translation system, lags behind others in reproducing genre-specific functions and ensuring comprehensibility. The study suggests that machine translation developers should leverage advanced AI language technology to develop an AI-assisted MT training system capable of interpreting nuances for more sophisticated translations.

In this paper, a new type of arm-type damper consisting of two links, two joints (universal joints with friction) and a hinge (a rotary oil damper) has been developed. The damper is able to suppress six modes of vibration of a machine, three in translational modes and three in rotational modes. The trial damper was made and the resisting force characteristics in the three translational directions were discussed theoretically and experimentally. The seismic responses of a three-degree-of-freedom system and the frequency response of a one-degree-of-freedom system, which consist of a mass and four coil springs and supported by the damper, were measured using a two-dimensional electro-hydraulic type shaking table. The experimental results are compared with the calculated ones, and the effects of vibration suppression of the damper are discussed.

With the development of artificial intelligence (AI), its application in literary translation has become more widespread, raising new possibilities and challenges for translating classical Chinese poetry. Against this backdrop, Qiangjinjiu is examined as one of the representative works of classical Chinese poetry, while relevant research from the perspective of Reception Theory remains scarce. Drawing on this theory, this paper compares AI and human translation of Qiangjinjiu from language, culture, and emotion. Findings show that although AI translation has advantages in terms of the information storage and translation output efficiency of poetry, it has significant deficiencies in conveying the imagery connotations, cultural metaphors, and rhythmic beauty of ancient poetry. Human translation, though less efficient, allows for deeper interpretation and more faithful restoration of the poems emotional and aesthetic depth, thereby better realizing its literary value. Based on this, this article believes that AI translation cannot completely replace human translation. In the future, it is necessary to promote the deep integration and coordinated development of the two, using AI to improve translation efficiency and human translation to optimize translation quality, ultimately facilitating the international dissemination and overseas inheritance of ancient poetry. The comparative study in this article can provide more powerful theoretical support for future research in the field of poetry translation, and at the same time guide translators to create poetry translations from the perspective of readers and take into greater consideration the degree of acceptance by readers.

Digitalization is one of the key distinctive features of modern environment and social life. Nowadays more and more functions are transferred to the artificial mind. How effective is the replacement of human activity with computer activity? In the given article, this problem is solved by an example of integration of digital technologies into translation activities. It this paper, emphasis is placed on the quality of machine translation (MT) output of legal texts in the language pair English - Slovak. It studies a Criminal Code formulated in the Slovak language which was translated by a human translator into English and consequently via machine translation system Google Translate (GT) back into Slovak. The back-translation - translation of a translated text back into its original language - as a quality assessment tool to detect discrepancies, mistranslations and inevitable differences between the source text and the target text was used. The quality of MT output was evaluated according to Multidimensional Quality Metrics (MQM) standards with the focus on the dimension of Fluency. The multiple comparisons were applied to determine which issues (errors) in Fluency dimension differ from the others. A statistically significant difference is noticed between Agreement and other issues, as well as between Ambiguity and other issues. The errors in Agreement are related to the differences between the languages: English is considered mostly an analytic language, Slovak represents a synthetic language. The issues in the Ambiguity dimension correlate with the type of the text being examined, since legal texts are characterized by relatively complicated wording and numerous terms; moreover, accuracy and unambiguity need to be preserved. Generally, the MT output is able to provide users with basic information about the text. On the other hand, most of the segments need revision and/or correction; in such cases, human intervention and post-editing is necessary.

: This memorandum report outlines the Human in the Loop Translation process, which strives to increase translators' productivity using statistical machine translation (SMT). In this process, U.S. Army Research Laboratory (ARL) systems trained by open source SMT toolkits were placed in a feedback loop with human translators in Afghanistan to incrementally improve Dari translations of English medical training manuals and to create bilingual glossaries in the medical domain. Anecdotal evidence indicates that the quality of the machine translation drafts produced using this feedback loop process is high enough to assist human translators in translating training manuals in the medical domain. Automatic scores showed large improvements in translation quality as the loop progressed for a very narrow task, providing indirect evidence that SMT can benefit human translators even with small amounts of training data.