{"id":"W2100039551","doi":"10.14778/1687627.1687735","title":"Distribution based microdata anonymization","year":2009,"lang":"en","type":"article","venue":"Proceedings of the VLDB Endowment","topic":"Privacy-Preserving Technologies in Data","field":"Computer Science","cited_by":11,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Toronto","funders":"","keywords":"Microdata (statistics); Computer science; Data anonymization; Closeness; Data mining; Aggregate (composite); Variety (cybernetics); Generalization; Information privacy; Artificial intelligence; Mathematics","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":["open_science"],"consensus_categories":["open_science"],"category_scores_codex":[0.0003161194,0.0001182891,0.0001162696,0.00005552923,0.0001078686,0.0001039368,0.01638636,0.00005934341,0.000004321179],"category_scores_gemma":[0.00368887,0.00008704625,0.00005321701,0.0005906645,0.0000561877,0.0005621918,0.01339389,0.0001183923,0.000005196266],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001115421,"about_ca_system_score_gemma":0.00002835729,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000007648882,"about_ca_topic_score_gemma":2.760306e-7,"domain_scores_codex":[0.9988794,0.000005650505,0.0002248936,0.0003269379,0.0003369921,0.0002261398],"domain_scores_gemma":[0.9977925,0.00001918221,0.0002204376,0.001794461,0.0001438523,0.00002954391],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.00002204924,0.0003569807,0.003106087,0.00006791879,0.00002203788,7.458037e-7,0.0000576077,0.00002182174,0.2138958,0.06580062,0.6724485,0.04419981],"study_design_scores_gemma":[0.0003991188,0.0001069295,0.006648937,0.00009299826,0.00001198652,0.000006012157,0.00001165848,0.03901171,0.7958396,0.1520302,0.00567097,0.0001698131],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.1160355,0.0003891682,0.5473781,0.3258309,0.001079521,0.001796729,0.0002204997,0.002036157,0.005233431],"genre_scores_gemma":[0.9016946,0.00002031977,0.0978747,0.0003291184,0.00002044693,0.00001163566,0.00001906094,0.000004252504,0.00002584191],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.7856591,"threshold_uncertainty_score":0.9945856,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01562681177843362,"score_gpt":0.2368745563587915,"score_spread":0.2212477445803579,"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."}}