{"id":"W4283008962","doi":"10.1371/journal.pone.0269097","title":"Measuring re-identification risk using a synthetic estimator to enable data sharing","year":2022,"lang":"en","type":"article","venue":"PLoS ONE","topic":"Privacy-Preserving Technologies in Data","field":"Computer Science","cited_by":21,"is_retracted":false,"has_abstract":true,"ca_institutions":"Children's Hospital of Eastern Ontario; University of Ottawa; University of Alberta","funders":"Natural Sciences and Engineering Research Council of Canada; Mitacs; FedDev Ontario; University of Toronto; Compute Canada","keywords":"Estimator; Computer science; Population; Statistics; Copula (linguistics); Data mining; Econometrics; Mathematics; Medicine","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":"metacan-v3-hybrid-931329e0061c","candidate_categories":["metaresearch"],"consensus_categories":[],"category_scores_codex":[0.0276988,0.0005863439,0.0005934798,0.001972165,0.0005336254,0.001826229,0.001321783,0.001319551,0.0008550179],"category_scores_gemma":[0.1175102,0.0003177556,0.0009281956,0.001594534,0.001011493,0.001991446,0.001919823,0.001351674,0.0002014927],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001294557,"about_ca_system_score_gemma":0.001324918,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002099908,"about_ca_topic_score_gemma":0.001490341,"domain_scores_codex":[0.984151,0.009942287,0.0009944377,0.001785089,0.002827111,0.0003001818],"domain_scores_gemma":[0.8550175,0.1004892,0.01399682,0.01975487,0.009877483,0.0008642545],"domain_codex":null,"domain_gemma":"reproducibility","domain_candidate":"reproducibility","domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0005764968,0.0004608993,0.1889412,0.0005943165,0.0009871341,0.0003112216,0.001111721,0.6290349,0.009094439,0.04096004,0.003275844,0.1246518],"study_design_scores_gemma":[0.00003827018,0.0006170265,0.0269948,0.0001210977,0.0001009518,0.0005847648,0.0003390484,0.9474185,0.009422516,0.01094117,0.003339863,0.00008194875],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.2707232,0.0004872181,0.7241307,0.0005438891,0.00009815208,0.0004259984,0.0008384426,0.0004747325,0.002277724],"genre_scores_gemma":[0.7910713,0.0001375712,0.2071885,0.0001176238,0.00003144717,0.0002459286,0.0008726037,0.00004154808,0.0002934737],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9723012,"threshold_uncertainty_score":0.146487,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.2833191170127614,"score_gpt":0.2998187574183161,"score_spread":0.01649964040555468,"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."}}