{"id":"W3021810331","doi":"10.1136/oem-2019-epi.106","title":"O4E.1 Cross-walking countries’ industry classifications using concordance files compared to fuzzy data matching, to utilise an international exposure dataset","year":2019,"lang":"en","type":"article","venue":"Occupational and Environmental Medicine","topic":"Nutritional Studies and Diet","field":"Medicine","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"","keywords":"Computer science; Data mining; Matching (statistics); Missing data; Statistics; Concordance; Artificial intelligence; Data science; Machine learning; Mathematics; Medicine","routes":{"ca_aff":false,"ca_fund":false,"ca_venue":false,"about_ca":true,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.007017747,0.001073721,0.0007172737,0.005698413,0.001202719,0.00426599,0.002206274,0.00116953,0.06607492],"category_scores_gemma":[0.04048263,0.0009246714,0.001800108,0.006455209,0.0006869585,0.003602221,0.004173576,0.0014327,0.02521736],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001502359,"about_ca_system_score_gemma":0.004093719,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.02448189,"about_ca_topic_score_gemma":0.0256535,"domain_scores_codex":[0.9962542,0.0005300643,0.0005467032,0.0009353243,0.001416229,0.0003174651],"domain_scores_gemma":[0.9848104,0.005202117,0.001047205,0.004315265,0.00418085,0.0004442625],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"observational","study_design_scores_codex":[0.001105678,0.0003514797,0.03402123,0.002316678,0.0002478016,0.0005718062,0.002855205,0.006074989,0.004359778,0.01855094,0.724452,0.2050925],"study_design_scores_gemma":[0.0004195294,0.0002694124,0.04598223,0.0009075681,0.00008330398,0.0003891068,0.002519227,0.04297636,0.01626575,0.01767819,0.8723285,0.0001808572],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"dataset","genre_gemma":"empirical","genre_scores_codex":[0.02166588,0.0001809441,0.1035749,0.001257771,0.0006324227,0.001646612,0.7071339,0.1375944,0.02631304],"genre_scores_gemma":[0.04992738,0.0001771334,0.2802565,0.0003201334,0.00008020314,0.001999373,0.6388349,0.01678996,0.01161441],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.06607492,"threshold_uncertainty_score":0.2210426,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.07674094737464159,"score_gpt":0.3708375650614487,"score_spread":0.2940966176868071,"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."}}