{"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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.02063317,0.0000813619,0.0001906795,0.0003083029,0.001035079,0.00008004743,0.001559801,0.00007028939,0.00003493906],"category_scores_gemma":[0.007080161,0.00007499472,0.00000664724,0.0002352349,0.00009139332,0.004650027,0.001093585,0.0002354954,0.000004858887],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0006694241,"about_ca_system_score_gemma":0.002363943,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0007425756,"about_ca_topic_score_gemma":0.0009310766,"domain_scores_codex":[0.9964536,0.0002428739,0.001529436,0.000370226,0.001105394,0.0002984177],"domain_scores_gemma":[0.9959214,0.0005353165,0.001961529,0.0006210578,0.0008077302,0.0001529954],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.0003392341,0.0001620649,0.8034119,0.0003929478,0.00006507834,0.000001145733,0.01561307,0.00004906923,0.0003531989,0.006272113,0.006378805,0.1669614],"study_design_scores_gemma":[0.001923973,0.00009227766,0.6588289,0.001292583,0.00001566358,0.00001386918,0.004820134,0.296625,0.0001048257,0.001257788,0.03477225,0.0002527981],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9599777,0.00003061399,0.03296404,0.002800529,0.001912508,0.0003904786,0.00177164,0.00001392341,0.0001385538],"genre_scores_gemma":[0.8189091,0.00004525279,0.1731632,0.0004896919,0.0005814524,0.00000382895,0.006760197,0.000006363809,0.00004097363],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.2965759,"threshold_uncertainty_score":0.8476124,"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."}}