{"id":"W3041938240","doi":"10.1109/msec.2020.2992821","title":"Seven Ways to Evaluate the Utility of Synthetic Data","year":2020,"lang":"en","type":"article","venue":"IEEE Security & Privacy","topic":"Privacy-Preserving Technologies in Data","field":"Computer Science","cited_by":64,"is_retracted":false,"has_abstract":true,"ca_institutions":"Children's Hospital of Eastern Ontario","funders":"","keywords":"Coronavirus disease 2019 (COVID-19); Data sharing; Computer science; Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2); Internet privacy; Medicine","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":[0.06236083,0.002094534,0.001044348,0.01250193,0.002027487,0.007622728,0.002202227,0.003719067,0.004373943],"category_scores_gemma":[0.2726549,0.0005105608,0.002286779,0.007821637,0.004702249,0.007026645,0.00429012,0.002868034,0.0007472197],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00407042,"about_ca_system_score_gemma":0.002530452,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.007316545,"about_ca_topic_score_gemma":0.005720499,"domain_scores_codex":[0.9495205,0.03285902,0.003751458,0.002160109,0.01049754,0.001211397],"domain_scores_gemma":[0.6837334,0.2528271,0.01014427,0.02146748,0.02848618,0.003341627],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[0.006676907,0.00172049,0.1796457,0.001868328,0.002262671,0.0009549781,0.002557734,0.3108652,0.004996517,0.1681433,0.01778249,0.3025256],"study_design_scores_gemma":[0.0004960801,0.002638991,0.02701242,0.001058553,0.0004559688,0.001010453,0.00552299,0.787684,0.01068897,0.1445106,0.01848775,0.0004333241],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"methods","genre_scores_codex":[0.5525708,0.005309097,0.366711,0.008980949,0.001790632,0.003330434,0.009547715,0.002259975,0.04949945],"genre_scores_gemma":[0.8239684,0.0006557777,0.1676647,0.0005992905,0.0001477718,0.0008691961,0.004120085,0.0002829218,0.001691887],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.06236083,"threshold_uncertainty_score":0.3297996,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1222913258340352,"score_gpt":0.3242382431896416,"score_spread":0.2019469173556064,"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."}}