{"id":"W4410447735","doi":"10.3233/shti250479","title":"Applying AI to Support Categorization of Heterogeneous Epidemiological Datasets","year":2025,"lang":"en","type":"article","venue":"Studies in health technology and informatics","topic":"Data Quality and Management","field":"Decision Sciences","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"McGill University Health Centre","funders":"Deutsche Forschungsgemeinschaft","keywords":"Computer science; Metadata; Annotation; Interoperability; World Wide Web; Information retrieval; Controlled vocabulary; Categorization; Data curation; Context (archaeology); Data science; Artificial intelligence","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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.01302614,0.001755202,0.001381379,0.0182983,0.002204288,0.004889603,0.003158175,0.001831324,0.002752013],"category_scores_gemma":[0.03940911,0.0004717422,0.002391349,0.0109921,0.001114998,0.005459531,0.003357831,0.002016199,0.002253486],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.003293274,"about_ca_system_score_gemma":0.004000057,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01729006,"about_ca_topic_score_gemma":0.01678904,"domain_scores_codex":[0.9891291,0.003450182,0.002149444,0.002381065,0.002537809,0.0003523382],"domain_scores_gemma":[0.9602622,0.02719111,0.001896825,0.00373437,0.006226179,0.000689305],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0004633514,0.0005716418,0.01880375,0.001831621,0.0004821614,0.0007154265,0.00279196,0.02016761,0.01680963,0.01585091,0.02119023,0.9003217],"study_design_scores_gemma":[0.000116942,0.0002133088,0.01150389,0.0005601853,0.0004178579,0.0005518321,0.002892487,0.7473791,0.02853584,0.1175321,0.09014,0.0001564999],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.02716581,0.001276187,0.9374896,0.002076978,0.0003159566,0.001301418,0.005598301,0.01960604,0.00516973],"genre_scores_gemma":[0.1172973,0.0005163137,0.8570922,0.000643163,0.0002191299,0.001016638,0.01998642,0.0005128941,0.002715918],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.0182983,"threshold_uncertainty_score":0.06888962,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.2480369898054427,"score_gpt":0.5283460737041448,"score_spread":0.2803090838987021,"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."}}