{"id":"W4402995867","doi":"10.1177/0282423x241266523","title":"Synthetic Census Microdata Generation: A Comparative Study of Synthesis Methods Examining the Trade-Off Between Disclosure Risk and Utility","year":2024,"lang":"en","type":"article","venue":"Journal of Official Statistics","topic":"Privacy-Preserving Technologies in Data","field":"Computer Science","cited_by":3,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"","keywords":"Microdata (statistics); Census; Econometrics; Computer science; Actuarial science; Geography; Business; Demography; Mathematics; Sociology; Population","routes":{"ca_aff":false,"ca_fund":false,"ca_venue":false,"about_ca":true,"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":[0.03256242,0.0007792209,0.0007567439,0.002944807,0.0005972725,0.002175373,0.001635604,0.001040118,0.001845987],"category_scores_gemma":[0.1516353,0.0004322714,0.0009954202,0.003174727,0.001496295,0.003284491,0.00264278,0.001303241,0.0004616716],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00178839,"about_ca_system_score_gemma":0.001765108,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003018341,"about_ca_topic_score_gemma":0.003176876,"domain_scores_codex":[0.9846922,0.0107105,0.0008429408,0.0009583084,0.002609783,0.000186276],"domain_scores_gemma":[0.8003849,0.1691533,0.005056656,0.01745133,0.007258001,0.0006958345],"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":[0.001673655,0.0004619028,0.03893105,0.002212061,0.0008095112,0.0002006922,0.001402457,0.4919114,0.002595124,0.06911457,0.01228983,0.3783978],"study_design_scores_gemma":[0.0002845979,0.0007843107,0.00775813,0.0005040807,0.0001425715,0.0003490703,0.001110456,0.891103,0.006805826,0.07256581,0.01848655,0.0001055045],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.3341689,0.008023031,0.6336961,0.003555309,0.0004236774,0.001687341,0.005913325,0.002329408,0.01020279],"genre_scores_gemma":[0.707714,0.002440165,0.2800271,0.0004301355,0.0001249397,0.0008893698,0.006777602,0.0004173559,0.001179323],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.03256242,"threshold_uncertainty_score":0.1722086,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1542671907916893,"score_gpt":0.3856461167268113,"score_spread":0.231378925935122,"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."}}