{"id":"W4391514282","doi":"10.48550/arxiv.2402.00179","title":"De-identification is not always enough","year":2024,"lang":"en","type":"preprint","venue":"arXiv (Cornell University)","topic":"Privacy-Preserving Technologies in Data","field":"Computer Science","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Natural Sciences and Engineering Research Council of Canada; University of Texas Health Science Center at Houston; National Institutes of Health; Cancer Prevention and Research Institute of Texas; National Science Foundation","keywords":"Identification (biology); Computer science; Biology","routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.02618423,0.000975009,0.001891937,0.00117245,0.002620489,0.006266724,0.002972474,0.004471053,0.005336598],"category_scores_gemma":[0.09134319,0.001049877,0.001327605,0.001929845,0.005086587,0.01453145,0.009117039,0.009330547,0.00449658],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001762061,"about_ca_system_score_gemma":0.003227053,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000680082,"about_ca_topic_score_gemma":0.0004924205,"domain_scores_codex":[0.9660763,0.01587641,0.00267617,0.005686481,0.008345473,0.001339196],"domain_scores_gemma":[0.8724129,0.04041702,0.006601333,0.07344081,0.005964254,0.001163585],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[0.00160558,0.0003611576,0.01222202,0.0008237389,0.0004462971,0.001208138,0.003312727,0.0457627,0.02073611,0.529256,0.02909569,0.3551698],"study_design_scores_gemma":[0.0001390261,0.0003593634,0.002457232,0.0005563435,0.0001307341,0.003643181,0.001483803,0.1415394,0.03838932,0.721966,0.08913349,0.0002021298],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":"theoretical_or_conceptual","genre_codex":"methods","genre_gemma":"commentary","genre_scores_codex":[0.04077082,0.001719425,0.92177,0.01702136,0.0008717371,0.0005126025,0.001894392,0.001819271,0.01362045],"genre_scores_gemma":[0.7446251,0.001352648,0.2350747,0.005645834,0.0006935675,0.0006525152,0.002843795,0.0004493178,0.008662706],"genre_candidate":"commentary","genre_consensus":null,"teacher_disagreement_score":0.02618423,"threshold_uncertainty_score":0.1384771,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.09957073959592029,"score_gpt":0.2193075750127301,"score_spread":0.1197368354168098,"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."}}