You will work on the ANSES project Unlocking the vaults of occupational exposure and health data (OPERAS).
Background: Large (occupational) epidemiological studies have been designed to have the power to detect exposure-response associations for a wide range of agents. High quality exposure assessment is crucial in such studies on occupational causes of diseases to estimate the prevalence and level of occupational exposure in the general population, to identify and quantify the potential risks associated with work circumstances and exposures, and to estimate the benefits in health and work participation from specific interventions and policies. The coding of each individual’s job history, an essential first step in the exposure assessment process, however, can be daunting and prohibitively time-consuming. As a result, many of the large studies are not maximally used for occupational health outcomes.
In OPERAS you will:
- develop algorithms to automatically code free-text entries of job titles, tasks, and industries into international standard classification of occupations (ISCO) and industries (ISIC/NACE);
- develop a Decision Support System (DSS): codes with low reliability will be flagged by the expert system and highlighted to assure a final check of a human coder; and
- validate the algorithms and the DST, using the French CONSTANCES cohort (200,000 entries) and the Dutch LIFEWORK cohort (25,000 entries).
In case you will be selected for the second round, an extensive project description will be provided.