{"id":"W2890617759","doi":"10.23889/ijpds.v3i4.945","title":"Development and validation of data quality rules in administrative health data using association rule mining","year":2018,"lang":"en","type":"article","venue":"International Journal for Population Data Science","topic":"Medical Coding and Health Information","field":"Health Professions","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"Canadian Institute for Health Information; University of Calgary","funders":"Canadian Institutes of Health Research","keywords":"Association rule learning; Data mining; Coding (social sciences); Data quality; Medicine; Computer science; Statistics; Mathematics; Operations management; Engineering","routes":{"ca_aff":true,"ca_fund":true,"ca_venue":false,"about_ca":true,"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":[0.1145951,0.001219735,0.001647407,0.0107164,0.00185729,0.006739826,0.003342402,0.00108449,0.001176367],"category_scores_gemma":[0.2475818,0.000652768,0.002688782,0.009209544,0.001632177,0.003180031,0.002975327,0.001909447,0.0005548132],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.003120455,"about_ca_system_score_gemma":0.007528659,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01052223,"about_ca_topic_score_gemma":0.009698235,"domain_scores_codex":[0.8853518,0.0540803,0.02484948,0.008971015,0.02508612,0.001661357],"domain_scores_gemma":[0.568929,0.3029788,0.04726542,0.02370572,0.05559225,0.001528851],"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":[0.0005433866,0.00107614,0.51911,0.005777718,0.0018509,0.0009780646,0.005719704,0.04191127,0.006112324,0.01226,0.008073683,0.3965869],"study_design_scores_gemma":[0.0003440204,0.001174203,0.2101044,0.006886106,0.001444174,0.001349216,0.009806329,0.633122,0.04604897,0.04376753,0.0455504,0.0004027101],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.3228067,0.001702047,0.645085,0.003434439,0.0002186077,0.006667933,0.01124029,0.002091354,0.006753598],"genre_scores_gemma":[0.2418468,0.0004693262,0.7459664,0.000326603,0.00005915322,0.001809149,0.009110903,0.00007026937,0.0003413818],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.1145951,"threshold_uncertainty_score":0.6060442,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.807174435393419,"score_gpt":0.661392854821811,"score_spread":0.1457815805716081,"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."}}