TEE4SOMC: Translation Equivalence Estimation for Statement of Merit Criteria
Bibliographic record
Abstract
This document reports on the results of Phase 1 of a research project to assess the feasibility of using Artificial Intelligence (AI) techniques to automatically evaluate translation quality in Statement of Merit Criteria (SOMC) data. This project is carried out by the National Research Council (NRC), in collaboration with the Public Service Commission of Canada (PSC). The PSC manages job advertisement for the Canadian government, and measures the overall compliance of job postings with official languages requirements. The main problem in this regard are ‘equivalence errors’ in SOMC’s, i.e. differences in the meaning of the English and French texts that can have an impact on applicants’ access to federal public jobs, as well as on the outcome of appointment processes. To support this project, PSC has provided NRC with a sample of SOMC data, in which ‘equivalence errors’ have been marked by PSC auditors. Analysis of this data reveals that about 10% of all SOMC sentences display equivalence issues. NRC proposed an AI method to assign quality scores to individual pairs of English-French sentences. Experiments on PSC data show that by using this method in an interactive setting, it would be possible to detect as much as 75% of all equivalence errors by controlling only 28% of the text. This method could be efficiently implemented as a software service for PSC, and various techniques exist that could further improve its performance.
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 | 0.013 | 0.094 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.009 | 0.005 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.003 | 0.002 |
| Open science | 0.002 | 0.002 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.011 | 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".