{"id":"W6958213133","doi":"10.6084/m9.figshare.15063757","title":"Additional file 2 of Assessing record linkage between health care and Vital Statistics databases using deterministic methods","year":2021,"lang":"en","type":"article","venue":"Figshare","topic":"Data Quality and Management","field":"Decision Sciences","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Institute of Health Economics; University of Calgary","funders":"","keywords":"Record linkage; Linkage (software); Health care; Data collection; MEDLINE; Data source; Health data","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":"codex-gemma-dda1882f352a","candidate_categories":["metaresearch","insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.0003364128,0.0001030543,0.0002958312,0.00009246239,0.0001858458,0.000265918,0.0002340661,0.00003134423,0.950251],"category_scores_gemma":[0.04089658,0.00009780863,0.00004124307,0.0002908104,0.000023375,0.000395103,0.0005905374,0.0001013544,0.0002240084],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0000350434,"about_ca_system_score_gemma":0.0003810091,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00001356091,"about_ca_topic_score_gemma":0.00005430214,"domain_scores_codex":[0.997871,0.0005165692,0.0005226538,0.0003849095,0.0005310574,0.0001737894],"domain_scores_gemma":[0.9824333,0.01629818,0.0004067336,0.0004135062,0.0003360587,0.000112193],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"not_applicable","study_design_scores_codex":[5.965148e-7,0.000007723255,0.00000666883,0.0001326435,0.000006861088,0.0000153459,0.00006362148,9.820675e-7,0.000001223372,0.00001260346,0.7497656,0.2499861],"study_design_scores_gemma":[0.0000621226,0.00003160624,0.002930671,0.001282188,0.000009401891,0.000005315949,0.001457143,0.000491303,0.00003338877,0.0003487351,0.9932339,0.0001142207],"study_design_candidate":"not_applicable","study_design_consensus":"not_applicable","genre_codex":"dataset","genre_gemma":"dataset","genre_scores_codex":[0.00001421656,0.0001066226,0.009701904,0.00002089461,0.00003890161,0.00007343734,0.989317,0.00001125335,0.0007157865],"genre_scores_gemma":[0.0001110801,9.413559e-7,0.3657191,0.0001405474,0.00005874089,0.00001761356,0.633797,0.000005356707,0.0001495705],"genre_candidate":"dataset","genre_consensus":"dataset","teacher_disagreement_score":0.950027,"threshold_uncertainty_score":0.9671823,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.5070898105060758,"score_gpt":0.552299466771692,"score_spread":0.04520965626561613,"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."}}