{"id":"W6958119743","doi":"10.6084/m9.figshare.15063754.v1","title":"Additional file 1 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":[0.0003383542,0.0001030888,0.000295755,0.00009252608,0.0001858314,0.0002653126,0.0002342031,0.00003134404,0.9511881],"category_scores_gemma":[0.04140037,0.00009784225,0.00004125146,0.0002910482,0.00002337534,0.0003948187,0.0005906248,0.000101379,0.0002265486],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00003503285,"about_ca_system_score_gemma":0.0003827995,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00001341869,"about_ca_topic_score_gemma":0.00005396626,"domain_scores_codex":[0.9978706,0.0005167504,0.0005224047,0.0003852109,0.0005311941,0.0001738558],"domain_scores_gemma":[0.9821886,0.01654174,0.0004061811,0.0004143882,0.0003368833,0.0001122227],"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":[6.030012e-7,0.000007765212,0.000006933762,0.0001320598,0.000006895926,0.00001499225,0.00006388848,9.821001e-7,0.000001276011,0.00001243735,0.7534129,0.2463393],"study_design_scores_gemma":[0.00006182364,0.00003167508,0.002988022,0.001302077,0.000009460052,0.000005316067,0.001476399,0.0004959442,0.0000336877,0.000346655,0.9931342,0.0001147429],"study_design_candidate":"not_applicable","study_design_consensus":"not_applicable","genre_codex":"dataset","genre_gemma":"dataset","genre_scores_codex":[0.00001431668,0.0001048501,0.009922705,0.00002100674,0.00003866084,0.0000735679,0.9891146,0.00001125512,0.0006989809],"genre_scores_gemma":[0.000112604,9.126549e-7,0.3681856,0.0001400376,0.00005763336,0.00001791162,0.6313314,0.000005345446,0.0001485213],"genre_candidate":"dataset","genre_consensus":"dataset","teacher_disagreement_score":0.9509616,"threshold_uncertainty_score":0.9666743,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.5172368392236222,"score_gpt":0.5531341020742365,"score_spread":0.03589726285061434,"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."}}