{"id":"W3208519060","doi":"10.1136/oem-2021-epi.80","title":"O-15 Occupational Health: A Multi-Cohort Job Title Cleaning Project by Algorithm","year":2021,"lang":"en","type":"article","venue":"Oral Presentations","topic":"Health, Environment, Cognitive Aging","field":"Environmental Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Dalhousie University","funders":"","keywords":"Computer science; Algorithm; Coding (social sciences); Cohort; General partnership; Cohort study; Medicine; Data mining; Finance; Statistics; Mathematics; Pathology","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":true,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.02486769,0.001340631,0.001285948,0.003147349,0.001873262,0.002539479,0.002985436,0.001345868,0.00968635],"category_scores_gemma":[0.05896249,0.0009721873,0.002569663,0.002653992,0.0008163875,0.001309035,0.002970938,0.001506111,0.002935086],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.003542133,"about_ca_system_score_gemma":0.01390716,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.07289049,"about_ca_topic_score_gemma":0.05959807,"domain_scores_codex":[0.9926234,0.00314114,0.0006230233,0.001728146,0.001543365,0.0003409514],"domain_scores_gemma":[0.9691532,0.01678079,0.0008600336,0.00314157,0.009557018,0.0005074609],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"observational","study_design_scores_codex":[0.001456401,0.0008998961,0.08822466,0.0009838572,0.0007027751,0.0002109768,0.001251421,0.07710689,0.002788521,0.009469011,0.05848978,0.7584159],"study_design_scores_gemma":[0.0009710105,0.0006152149,0.02217159,0.0003524697,0.000411047,0.0002830594,0.001207725,0.9136884,0.004820294,0.0154386,0.03995579,0.00008481528],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.08691274,0.0005093162,0.880769,0.00144755,0.0002773145,0.006043609,0.007731903,0.01078993,0.0055187],"genre_scores_gemma":[0.06406973,0.0001133964,0.9217134,0.0002204449,0.00002599677,0.002501642,0.008090268,0.0006758466,0.002589356],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.07289049,"threshold_uncertainty_score":0.1449324,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.05824154844107707,"score_gpt":0.3672206555606273,"score_spread":0.3089791071195502,"validation_status":"score_only:v0-immature-baseline","note":"Baseline scores from an immature model (maturity gate not passed). Scores rank; they never assert a category."}}