{"id":"W2934046129","doi":"10.2478/jos-2019-0007","title":"An Evolutionary Schema for Using “it-is-what-it-is” Data in Official Statistics","year":2019,"lang":"en","type":"article","venue":"Journal of Official Statistics","topic":"Data Quality and Management","field":"Decision Sciences","cited_by":7,"is_retracted":false,"has_abstract":true,"ca_institutions":"Health Canada","funders":"","keywords":"Data science; Schema (genetic algorithms); Computer science; Set (abstract data type); Big data; Agency (philosophy); Process (computing); Data mining; Information retrieval; Sociology; Social science","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":[0.09347548,0.0007567634,0.001092298,0.009199608,0.006293026,0.02111887,0.005563634,0.006896842,0.003956242],"category_scores_gemma":[0.07339408,0.001335479,0.002209512,0.0118202,0.02743293,0.03810327,0.0104171,0.01041549,0.001520624],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.007187177,"about_ca_system_score_gemma":0.01124555,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.006836001,"about_ca_topic_score_gemma":0.005993392,"domain_scores_codex":[0.9177893,0.05679551,0.008791584,0.00553027,0.009553523,0.00153983],"domain_scores_gemma":[0.9128823,0.03841355,0.005530451,0.01746239,0.02182569,0.0038857],"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":[0.000008280497,0.00002179152,0.000649076,0.00009297198,0.00001875795,0.00008600347,0.005936255,0.0003549281,0.0002122302,0.9784362,0.003546352,0.0106372],"study_design_scores_gemma":[0.00002718132,0.00006763402,0.0007998889,0.000861331,0.00005835874,0.0006189221,0.01048875,0.004717417,0.001071321,0.6268383,0.3543591,0.00009186374],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":"theoretical_or_conceptual","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.00590991,0.001015369,0.9264815,0.04166785,0.001102299,0.0007427885,0.0005631027,0.0008078174,0.02170944],"genre_scores_gemma":[0.0673285,0.0008418838,0.9229077,0.003815326,0.0002784835,0.001022775,0.0007298996,0.0003245979,0.002750861],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.09347548,"threshold_uncertainty_score":0.4943514,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.3369684994523318,"score_gpt":0.4966412882000322,"score_spread":0.1596727887477005,"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."}}