Six NMT Systems, One Language Pair: Which Best Translates Arabic-English?
Bibliographic record
Abstract
This study evaluates the quality of translations produced by six different Neural Machine Translation (NMT) systems when translating from Arabic to English. The systems under study are Google Translate, Microsoft Bing, Yandex, Systran, ChatGPT-4, and Amazon Translate. Given the precision and complexity of the Arabic language, the study aims to examine the most effective NMT system and understand how translators can utilize these tools. To achieve the study's objectives, 1,000 Arabic sentences and their English translations are examined, with established translations verified by human translators used as reference benchmarks for evaluating machine translations. Data are collected from the Tatoeba platform (2024), accessible to researchers online, and are analyzed using the Bilingual Evaluation Understudy BLEU system. The study's findings reveal significant variations in translation quality among the systems tested, highlighting the necessity for translators to be involved in the machine translation editing process. Moreover, the results indicate that ChatGPT-4 outperform other systems in producing high-quality translations. This study contributes to translation studies by offering a comprehensive comparative analysis of current NMT systems, providing practical insights for translators, and advancing research on machine translation applications.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.004 | 0.021 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.002 | 0.001 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.002 | 0.002 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.003 | 0.003 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".