{"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":"codex-gemma-dda1882f352a","candidate_categories":["insufficient_payload"],"consensus_categories":["insufficient_payload"],"category_scores_codex":[0.000179709,0.00007383388,0.00007804651,0.00002089658,0.0002084873,0.00002691952,0.00006811813,0.00002944931,0.005930095],"category_scores_gemma":[0.00005570647,0.00008020129,0.00002911713,0.0001800684,0.00006129128,0.0001602509,0.00009161636,0.0001161866,0.002468263],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000122662,"about_ca_system_score_gemma":0.00007187785,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001063842,"about_ca_topic_score_gemma":0.0002039715,"domain_scores_codex":[0.9989844,0.0001120173,0.0001498592,0.0002860487,0.0002380358,0.0002296269],"domain_scores_gemma":[0.9996726,0.00004376673,0.00004768349,0.0001456336,0.000007913875,0.00008243387],"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":[0.000004095968,0.0003278231,0.3922069,0.00001409207,0.00003941418,0.00002354921,0.0008732947,0.0002127025,0.001719815,0.0001753467,0.4778591,0.1265438],"study_design_scores_gemma":[0.0005783875,0.00003268834,0.68381,0.00001477087,0.00001788711,0.00001672392,0.0006483769,0.010737,0.000846436,0.0001465971,0.3028674,0.0002837167],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.1354463,0.0008671673,0.4443133,0.008955909,0.001195282,0.003691664,0.001103438,0.000483998,0.403943],"genre_scores_gemma":[0.5070198,0.0004556636,0.3843745,0.009754022,0.0003330927,0.0004548384,0.002068291,0.0001478371,0.09539195],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.3715736,"threshold_uncertainty_score":0.9983084,"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."}}