{"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":"codex-gemma-dda1882f352a","candidate_categories":["metaresearch","scholarly_communication"],"consensus_categories":["metaresearch"],"category_scores_codex":[0.06985871,0.0000944163,0.0001830788,0.0007846038,0.0008914953,0.003587579,0.003888494,0.00001455246,0.00004774836],"category_scores_gemma":[0.03550597,0.0000609864,0.00003208081,0.001023515,0.000238221,0.005406113,0.002240151,0.0002666804,0.000008307293],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001678084,"about_ca_system_score_gemma":0.0004475631,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001388872,"about_ca_topic_score_gemma":0.0006335622,"domain_scores_codex":[0.9932593,0.0002801639,0.001757249,0.000783774,0.003620058,0.0002994516],"domain_scores_gemma":[0.993659,0.002849753,0.000973883,0.001153176,0.00112775,0.000236404],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"observational","study_design_scores_codex":[0.0001916164,0.0002138101,0.06611174,0.00008625077,0.00009549926,0.00003688538,0.009235814,0.0005274934,0.0005458561,0.1494339,0.05881582,0.7147053],"study_design_scores_gemma":[0.0003920547,0.001503763,0.4259354,0.001620715,0.0000146492,0.0000553623,0.01062984,0.04116329,0.0006195438,0.1123572,0.4053451,0.0003631605],"study_design_candidate":"design_other","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.8729688,0.00006272398,0.0783184,0.04249705,0.00155874,0.0008515028,0.003351486,0.00002632206,0.0003650106],"genre_scores_gemma":[0.9868198,0.00002677111,0.01188766,0.0002607693,0.0002495353,0.00001148999,0.0002890789,0.000005856443,0.0004490145],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.7143422,"threshold_uncertainty_score":0.9974468,"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."}}