{"id":"W3209399758","doi":"10.1136/oem-2021-epi.442","title":"S-464 Automated Occupational Encoding to the Canadian National Occupation Classification using an Ensemble Classifier from TF-IDF and Doc2Vec Embeddings","year":2021,"lang":"en","type":"article","venue":"","topic":"Artificial Intelligence in Healthcare","field":"Health Professions","cited_by":3,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of New Brunswick","funders":"","keywords":"Computer science; Coding (social sciences); Classifier (UML); Support vector machine; Data mining; Artificial intelligence; Machine learning","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":["sts","insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.0009913504,0.0001929174,0.0002022252,0.000220944,0.003110151,0.0001255511,0.000197944,0.0003504364,0.001919126],"category_scores_gemma":[0.001536702,0.0001520805,0.00003412754,0.0006472865,0.00006544905,0.0004696189,0.00008281141,0.000493784,0.0003453191],"about_ca_system_candidate":true,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001480088,"about_ca_system_score_gemma":0.00612927,"about_ca_topic_candidate":true,"about_ca_topic_consensus":true,"about_ca_topic_score_codex":0.3326193,"about_ca_topic_score_gemma":0.9203997,"domain_scores_codex":[0.996809,0.0006924493,0.000760717,0.0005652804,0.0006086801,0.0005638628],"domain_scores_gemma":[0.9963114,0.0009817695,0.0002181029,0.0003472388,0.00163926,0.0005022755],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.000395005,0.0001641465,0.7985556,0.0001393947,0.00009987915,0.00003690206,0.03623619,0.003521952,0.02227304,0.08049498,0.04481579,0.01326717],"study_design_scores_gemma":[0.0001542099,0.00003763334,0.2928399,0.0001608545,0.000023073,0.000008160879,0.01565657,0.6714233,0.00091294,0.00231803,0.01614825,0.0003170864],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9752223,0.00003698769,0.004500506,0.0100822,0.001148016,0.001002434,0.0002502153,0.000300738,0.00745655],"genre_scores_gemma":[0.9834871,0.000009353929,0.007179814,0.007258827,0.0006494774,0.0001175903,0.0004685677,0.00003046535,0.0007988046],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.6679013,"threshold_uncertainty_score":0.999505,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.3473811814956081,"score_gpt":0.5158195507423838,"score_spread":0.1684383692467757,"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."}}