Developing Post-editing Strategies for Machine-translated Chinese Texts: A Case Study on Gynaecological Cancer Information
Notice bibliographique
Résumé
Concomitant with the increased development of the Internet, the number of users in China has increased, and more people now use it to access health information. The incidence of gynaecological cancer (GC) in China is on the rise. However, the online availability of information concerning GC in China is limited compared with that available in the United Kingdom, the United States of America, Australia, Canada, and New Zealand. The lack of this information limits people’s awareness of healthcare. As deep-learning techniques have developed, neural machine translation (NMT) has become a mainstream approach in machine translation (MT). Free-to-use tools offer a low-cost and highly efficient language-translation solution. However, despite continued advancement in NMT technology in processing texts from English to Chinese, errors persist in scientific and technical texts, especially those of a medical nature. Even when accounting for people’s socio-economic status, language barriers that limit communication can significantly affect health. By incorporating human post-editing, MT could reduce the time and cost required to translate public health materials from English to Chinese that maintain a similar quality to human translation. By doing so, the availability of multilingual public health materials would be significantly improved. Inaccurate, ambiguous, unnatural, or non-inclusive use of translated language may generate misunderstanding regarding information in translations. In medical texts, such as those pertaining to GCs, this could lead to inappropriate decisions being made, and even negative impacts on psychological or mental health. Therefore, the objectives of research presented herein are to develop post-editing strategies to deal with machine translations from English to Chinese of medical texts, and specifically texts pertaining to GC, based on language use guides for cancer information. In doing so, the accuracy, clarity, and naturalness of health information is improved, and that information available to Chinese-speaking people with cancer, and their families and friends, would be more positive and supportive, and public health awareness would be improved. To achieve these objectives, online health information was sourced from health departments and organisations in the United States of America, United Kingdom, Canada, Australia, and New Zealand (e.g., the National Cancer Institute, Cancer Research UK, Canadian Cancer Society, Cancer Council Australia, and The Cancer Society of New Zealand). By way of qualitative analysis, the limitations of English to Chinese machine-translation of GC information (accuracy at lexical and syntactic levels, logical coherence, lexical and syntactic ambiguities, idiomatic expression, and linguistic inclusiveness) are evaluated. New post-editing strategies of machine-translated texts are developed, and their ability to resolve various MT issues is demonstrated in a series of case studies.
Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.
Comment cette classification a été obtenuedéplier
Prédiction machine sur la base complète
Imitation des enseignantsNi prévalence calibrée, ni vérité terrain. Validation humaine à venir. Le volet Gemma est une étiquette directe du modèle pour chaque travail de la base, lue sur la notice réduite au titre. Le volet Codex est un classifieur appris des 10 348 étiquettes directes de Codex et calibré sur les taux pondérés de l'échantillon; les champs sans appui suffisant ne portent aucun appel Codex. Le mode candidate est l'union des deux volets; le consensus est leur intersection. Ces sorties portent le statut machine_predicted_unvalidated et ne sont pas des étiquettes humaines.
Scores du classifieur distillé par catégorie (deux têtes)
| Catégorie | Codex | Gemma |
|---|---|---|
| Métarecherche | 0,004 | 0,027 |
| Méta-épidémiologie (sens strict) | 0,001 | 0,000 |
| Méta-épidémiologie (sens large) | 0,001 | 0,001 |
| Bibliométrie | 0,002 | 0,002 |
| Études des sciences et des technologies | 0,003 | 0,001 |
| Communication savante | 0,002 | 0,004 |
| Science ouverte | 0,002 | 0,002 |
| Intégrité de la recherche | 0,002 | 0,002 |
| Charge utile insuffisante (le modèle a refusé de juger) | 0,006 | 0,003 |
Scores machine (provisoires)
Les deux têtes enseignantes du modèle étudiant, lues sur ce travail. Un score ordonne la base pour la relecture; il n'affirme jamais une catégorie, et le statut de validation accompagne chaque rangée tel quel.
Scores de référence d'un modèle non mature (critères de maturité non atteints, 7 itérations). Un score ordonne; il n'affirme jamais une catégorie.
score_only:v0-immature-baseline · tel quel depuis la passe de notation : score_only signifie que le nombre peut ordonner les travaux, et qu'aucune étiquette de catégorie n'en découleClassification
machine, non validéePrédiction automatique; un appel candidat d’une seule source (Gemma direct ou Codex distillé), pas un consensus.
Le détail, modèle par modèle et score par score, se trouve en fin de page sous « Comment cette classification a été obtenue ».