{"id":"W2341199921","doi":"10.1136/oemed-2015-103152","title":"Computer-based coding of free-text job descriptions to efficiently identify occupations in epidemiological studies","year":2016,"lang":"en","type":"article","venue":"Occupational and Environmental Medicine","topic":"Occupational and environmental lung diseases","field":"Medicine","cited_by":62,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Alberta","funders":"National Institutes of Health","keywords":"Computer science; Coding (social sciences); Terminology; Artificial intelligence; Machine learning; Data mining; Statistics; Mathematics","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0002557076,0.000200131,0.000428555,0.0002286382,0.00009268352,0.000002627969,0.00009143777,0.00006207423,0.0004333304],"category_scores_gemma":[0.0003707278,0.0001226203,0.00007127713,0.0001546032,0.0005975116,0.00008752349,0.000117446,0.00007211448,0.00003925535],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000205293,"about_ca_system_score_gemma":0.0000214615,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00003742679,"about_ca_topic_score_gemma":0.00001361012,"domain_scores_codex":[0.9983959,0.00006898052,0.0005344881,0.0003638775,0.0004223166,0.0002144725],"domain_scores_gemma":[0.9988298,0.0005941552,0.0001035698,0.0002234831,0.00001625687,0.0002327333],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.0005278461,0.0003868485,0.9691945,0.00006103538,0.00004621702,0.00001147129,0.0001066369,0.0002730228,0.02132306,0.0004951035,0.001666105,0.005908113],"study_design_scores_gemma":[0.002435221,0.0006994088,0.9941599,0.0006769868,0.00007376099,0.0000132535,0.0001948803,0.0004499474,0.0003923911,0.0003662101,0.0004027826,0.0001352083],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9902027,0.0006677799,0.004385598,0.003832636,0.0001315594,0.0003700218,0.0002603043,0.00001993952,0.0001294522],"genre_scores_gemma":[0.9951009,0.0002949669,0.00278897,0.001304626,0.0001402826,0.00005921837,0.0001564863,0.00001145341,0.0001431577],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.02496541,"threshold_uncertainty_score":0.5000313,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.06200902610992798,"score_gpt":0.3360384187649715,"score_spread":0.2740293926550436,"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."}}