{"id":"W4402405684","doi":"10.23889/ijpds.v9i5.2760","title":"Reporting on the establishment of a Privacy Preserving Record Linkage to Facilitate an Ongoing Crosswalk Between Research and Health Administrative Databases","year":2024,"lang":"en","type":"article","venue":"International Journal for Population Data Science","topic":"Data Quality and Management","field":"Decision Sciences","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Indoc Research; Ontario Brain Institute","funders":"","keywords":"Schema crosswalk; Record linkage; Linkage (software); Database; Data science; Computer science; Internet privacy; Data mining; Business; Medicine; Engineering; Environmental health; Transport engineering","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":true,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[{"model":"gemma","categories":[],"domain":null,"study_design":"not_applicable","genre":"methods","about_ca_system":false,"about_ca_topic":false,"confidence":"low","status":"direct model label, unvalidated"},{"model":"gpt","categories":[],"domain":null,"study_design":"design_other","genre":"methods","about_ca_system":false,"about_ca_topic":false,"confidence":"low","status":"direct model label, unvalidated"}],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":["metaresearch"],"consensus_categories":[],"category_scores_codex":[0.3317976,0.001554459,0.001456643,0.008349538,0.006821468,0.01009345,0.005997896,0.004587407,0.02258424],"category_scores_gemma":[0.4838728,0.001968247,0.003311297,0.00983963,0.003647804,0.008550133,0.01760065,0.005553245,0.01024992],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.009969041,"about_ca_system_score_gemma":0.06828897,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01306514,"about_ca_topic_score_gemma":0.01256952,"domain_scores_codex":[0.5684075,0.2431039,0.07261993,0.01667148,0.0900703,0.009126974],"domain_scores_gemma":[0.3587295,0.1729114,0.08027948,0.1830641,0.1952448,0.009770704],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"not_applicable","study_design_scores_codex":[0.001571361,0.0007277385,0.05017734,0.007127367,0.0007511079,0.001046822,0.01560221,0.003450872,0.006395472,0.09431692,0.2901109,0.5287219],"study_design_scores_gemma":[0.0004537326,0.001049532,0.02958041,0.007381015,0.0004842972,0.0009450627,0.006405848,0.005178116,0.02130152,0.03129448,0.8955286,0.0003973305],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.03993421,0.002617937,0.6750861,0.07353607,0.007190322,0.05455412,0.05013268,0.006518905,0.09042959],"genre_scores_gemma":[0.1442952,0.003442467,0.7001932,0.01365849,0.00254804,0.06772459,0.0389197,0.001351809,0.02786656],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.3317976,"threshold_uncertainty_score":0.8240125,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.9109178340257105,"score_gpt":0.6727323830946649,"score_spread":0.2381854509310456,"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."}}