Development of and Selected Performance Characteristics of CANJEM, a General Population Job-Exposure Matrix Based on Past Expert Assessments of Exposure
Bibliographic record
Abstract
Objectives: We developed a job-exposure matrix called CANJEM using data generated in population-based case-control studies of cancer. This article describes some of the decisions in developing CANJEM, and some of its performance characteristics. Methods: CANJEM is built from exposure information from 31673 jobs held by study subjects included in our past case-control studies. For each job, experts had evaluated the intensity, frequency, and likelihood of exposure to a predefined list of agents based on jobs histories and descriptions of tasks and workplaces. The creation of CANJEM involved a host of decisions regarding the structure of CANJEM, and operational decisions regarding which parameters to present. The goal was to produce an instrument that would provide great flexibility to the user. In addition to describing these decisions, we conducted analyses to assess how well CANJEM covered the range of occupations found in Canada. Results: Even at quite a high level of resolution of the occupation classifications and time periods, over 90% of the recent Canadian working population would be covered by CANJEM. Prevalence of exposure of specific agents in specific occupations ranges from 0% to nearly 100%, thereby providing the user with basic information to discriminate exposed from unexposed workers. Furthermore, among exposed workers there is information that can be used to discriminate those with high exposure from those with low exposure. Conclusions: CANJEM provides good coverage of the Canadian working population and possibly that of several other countries. Available in several occupation classification systems and including 258 agents, CANJEM can be used to support exposure assessment efforts in epidemiology and prevention of occupational diseases.
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 | 0.007 | 0.029 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.005 | 0.003 |
| Science and technology studies | 0.001 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.003 | 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; 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".