{"id":"W3196335171","doi":"10.20517/jsss.2020.20","title":"Information theoretic-based privacy risk evaluation for data anonymization","year":2022,"lang":"en","type":"article","venue":"Journal of Surveillance Security and Safety","topic":"Privacy-Preserving Technologies in Data","field":"Computer Science","cited_by":7,"is_retracted":false,"has_abstract":true,"ca_institutions":"Polytechnique Montréal","funders":"","keywords":"Metric (unit); Computer science; Inference; Data anonymization; Data publishing; Identification (biology); Data mining; Process (computing); Information sensitivity; Information privacy; Publishing; Artificial intelligence; Computer security","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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.03772826,0.00226798,0.001799716,0.007602811,0.001561731,0.005780879,0.002707756,0.002987427,0.001574175],"category_scores_gemma":[0.138918,0.0006077428,0.001970136,0.00498546,0.006023956,0.01046312,0.005550467,0.003389322,0.0003045294],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.006191717,"about_ca_system_score_gemma":0.003704793,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0009702196,"about_ca_topic_score_gemma":0.0006142301,"domain_scores_codex":[0.9245122,0.04235813,0.004052348,0.004526381,0.02331181,0.00123914],"domain_scores_gemma":[0.8293856,0.1205103,0.01575317,0.02047915,0.01155197,0.002319861],"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":[0.0007941486,0.0003385174,0.02062572,0.001098321,0.0009026644,0.0003344842,0.0006756606,0.4991122,0.003239299,0.3474324,0.003976131,0.1214705],"study_design_scores_gemma":[0.00004562557,0.001112703,0.005508541,0.000346162,0.0002547424,0.001154085,0.0003904924,0.7424431,0.00601357,0.2357033,0.006826689,0.000200886],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.03798229,0.00296997,0.9499437,0.001403172,0.0001615572,0.0004302674,0.0006756868,0.0002550301,0.006178318],"genre_scores_gemma":[0.821171,0.001365215,0.1741149,0.0003963422,0.000282516,0.0006014473,0.0009930168,0.00009167282,0.000983783],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.03772826,"threshold_uncertainty_score":0.1995285,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02722922051687764,"score_gpt":0.2842230224223194,"score_spread":0.2569938019054417,"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."}}