{"id":"W4323349011","doi":"10.56553/popets-2023-0055","title":"A Unified Framework for Quantifying Privacy Risk in Synthetic Data","year":2023,"lang":"en","type":"article","venue":"Proceedings on Privacy Enhancing Technologies","topic":"Privacy-Preserving Technologies in Data","field":"Computer Science","cited_by":72,"is_retracted":false,"has_abstract":true,"ca_institutions":"Vector Institute","funders":"","keywords":"Computer science; Differential privacy; Information privacy; Data anonymization; Data mining; Inference; Privacy software; Set (abstract data type); Information sensitivity; Computer security; Synthetic data; Privacy protection; Data science; Algorithm; Artificial intelligence","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":"codex-gemma-dda1882f352a","candidate_categories":["metaresearch","metaepi_narrow","open_science"],"consensus_categories":["open_science"],"category_scores_codex":[0.003549858,0.0006327854,0.0007243357,0.002053457,0.0006518154,0.0009408172,0.08080528,0.0009013638,0.000002710998],"category_scores_gemma":[0.3748322,0.0006108578,0.0001134041,0.005412232,0.0003590478,0.002312512,0.1696805,0.001836878,0.0001409593],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000349236,"about_ca_system_score_gemma":0.0001485766,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00004711534,"about_ca_topic_score_gemma":0.00002925507,"domain_scores_codex":[0.993901,0.00003253987,0.0009575443,0.002706242,0.0007921466,0.001610506],"domain_scores_gemma":[0.9806677,0.002190949,0.0007262735,0.01608202,0.0002570979,0.00007602697],"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":[0.0001598481,0.0004480337,0.0126792,0.00108613,0.0001821475,0.00007411444,0.001525953,0.00003458793,0.01140708,0.5422849,0.09814738,0.3319707],"study_design_scores_gemma":[0.0004056773,0.0001819556,0.0006412371,0.0008183802,0.00001632373,0.00001007471,0.0006397642,0.1445733,0.06927429,0.7799497,0.002852897,0.0006364192],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":"theoretical_or_conceptual","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.2958537,0.0004192872,0.5746897,0.07648461,0.0009383871,0.002501614,0.0001245041,0.04863667,0.0003514682],"genre_scores_gemma":[0.5388239,0.0005216227,0.4601712,0.00005160348,0.00002544523,0.0003252558,0.0000182359,0.00005324454,0.000009521907],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.3712823,"threshold_uncertainty_score":0.9996343,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1012695659724597,"score_gpt":0.3424893920809355,"score_spread":0.2412198261084758,"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."}}