{"id":"W2146453464","doi":"10.1186/1472-6947-12-66","title":"Estimating the re-identification risk of clinical data sets","year":2012,"lang":"en","type":"article","venue":"BMC Medical Informatics and Decision Making","topic":"Privacy-Preserving Technologies in Data","field":"Computer Science","cited_by":102,"is_retracted":false,"has_abstract":true,"ca_institutions":"Agricultural Research Institute of Ontario; University of Ottawa; Children's Hospital of Eastern Ontario","funders":"Canada Research Chairs","keywords":"Estimator; Statistics; Uniqueness; Quartile; Identification (biology); Mathematics; Data mining; Computer science; Econometrics; Confidence interval","routes":{"ca_aff":true,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.09864182,0.0009115613,0.001737957,0.003322722,0.0009672338,0.002495436,0.002013796,0.002030286,0.0004819758],"category_scores_gemma":[0.3289839,0.0006411585,0.001642058,0.001468769,0.002064965,0.003150439,0.002263698,0.002825836,0.0001231872],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00199373,"about_ca_system_score_gemma":0.002451013,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002031152,"about_ca_topic_score_gemma":0.001413561,"domain_scores_codex":[0.9501701,0.03182235,0.004678455,0.005592984,0.006978034,0.0007580619],"domain_scores_gemma":[0.4139811,0.5407919,0.02250531,0.01244479,0.009270987,0.001005893],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00149104,0.0003421092,0.2770668,0.00074732,0.001566767,0.0004723874,0.001585374,0.4898333,0.003224698,0.01689885,0.001377864,0.2053936],"study_design_scores_gemma":[0.0001020495,0.0005866492,0.02216746,0.0002527453,0.0002279766,0.0005879872,0.0003106609,0.9456778,0.008396224,0.02084311,0.0007414986,0.0001059128],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.5297948,0.00128509,0.4657547,0.0009191771,0.00007081774,0.0006095438,0.0002894883,0.000331895,0.0009445148],"genre_scores_gemma":[0.8080102,0.0002122646,0.1907671,0.0001551725,0.00003624627,0.0002848183,0.0003301664,0.00002938399,0.0001746305],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.09864182,"threshold_uncertainty_score":0.521674,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1848199606233114,"score_gpt":0.4447147731322799,"score_spread":0.2598948125089684,"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."}}