Analyse comparative de la couverture et de l’acceptabilité des solutions d’un dictionnaire bilingue spécialisé, d’une banque de données terminologiques et d’un concordancier en ligne : application au domaine de la traduction médicale
Bibliographic record
Abstract
La présente étude rend compte de l’utilité de trois ressources pour le traducteur médical, soit le Dictionnaire anglais-français des sciences médicales et paramédicales de Gladstone, la banque de données terminologiques et linguistiques du gouvernement du Canada TERMIUM Plus® et le concordancier en ligne WeBiText, sur la base d’une analyse comparative de leur taux de couverture et du niveau d’acceptabilité des solutions qu’elles proposent. Contrairement à d’autres études, cette recherche mise sur une approche double qui combine un volet quantitatif et qualitatif. À partir d’un ensemble de problèmes de traduction tiré d’un corpus d’étude constitué de huit résumés d’articles scientifiques de revues médicales canadiennes, nous démontrons que le taux de couverture du corpus « Santé » de WeBiText est de 62 %, alors que celui de TERMIUM Plus® est de 37 % et celui du Gladstone de 30 %. Nous montrons par ailleurs que le concordancier fournit environ deux fois plus de solutions que les deux autres ressources. En outre, des données rassemblées dans le cadre d’un questionnaire en ligne auprès de 88 langagiers révèlent que les solutions proposées par WeBiText sont tout aussi acceptables que celles offertes par TERMIUM Plus®, tandis que celles qui figurent dans le Gladstone sont souvent considérées comme inacceptables. Nous en concluons que les corpus parallèles tirés du Web ont une meilleure couverture que les ressources traditionnelles et qu’ils présentent une gamme plus variée de solutions uniques et convenables, notamment pour la traduction de textes médicaux hautement spécialisés. ENGLISH This study reports on the usefulness of three resources for the medical translator, namely Gladstone's English-French dictionary of medical and paramedical sciences, TERMIUM Plus®, the Government of Canada's terminology and linguistic data bank, and WeBiText, an online concordancer, on the basis of a comparative analysis of their coverage and the level of acceptability of the solutions they offer. Unlike other studies, this research adopts a two-fold approach combining quantitative and qualitative perspectives. From a set of translation problems drawn from a corpus of eight scientific paper abstracts from Canadian medical journals, we demonstrate that WeBiText's "Health" corpus has a 62% coverage rate, while that of TERMIUM Plus® is 37% and Gladstone's is 30%. We also show that the concordancer provides roughly twice as many solutions as the other two resources. In addition, data collected through an online survey of 88 language professionals indicates that the solutions found in WeBiText are equally acceptable as those provided by TERMIUM Plus®, while Gladstone's are often considered unacceptable. We conclude that Web parallel corpora have a better coverage than traditional resources and that they offer a broader range of unique and suitable solutions, particularly for the translation of highly specialized medical texts.
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 distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.004 | 0.001 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.000 |
| Bibliometrics | 0.001 | 0.000 |
| Science and technology studies | 0.004 | 0.007 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.001 | 0.000 |
| Research integrity | 0.001 | 0.002 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
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; both teacher heads agree on what is shown here.
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".